After this post goes up, I’m going to put Singal-Minded to sleep for roughly a month and a half. If you’re a paying subscriber, you don’t have to do anything — you won’t get charged while the newsletter is dormant. (I am finishing up one piece for The Dispatch that you will receive, as always, a day after it goes up.)

I’m partly doing this because I’m entering (another) period of book crunch and need to devote as much of my brain as possible to that during a chunk of this summer. I’m quite excited for this book but I really, really need to finish it so I can move on.

That’s not my only motivation, though. A pause like this also provides an opportunity to reboot Singal-Minded, and to improve its quality.

I’ve been writing this newsletter since 2019. I’m proud of Singal-Minded, not to mention eternally grateful that it has built up the audience it has. But, in large part because my attention is perpetually divided between this and other projects, it’s not as strong a newsletter as it could be.

I plan to reboot Singal-Minded with a slightly different focus, with a more consistent publication schedule, and with other features that will, I think, help it to stand out in an ever-more-competitive Substack landscape. I’ve already taken a first step on this front: I’ve hired an assistant editor who I will formally introduce when I return. He and I have been brainstorming how to make Singal-Minded 2.0 as strong as possible, and he’s going to be doing a lot of tinkering between now and then. At some point before the relaunch, I’m also going to solicit input from my subscribers about what they want to see (or don’t want to see) in this newsletter going forward. So please keep an eye out for that email.

In the meantime, I’m curious to hear any off-the-cuff thoughts you might have about this here publication. Paid and free subscribers alike are welcome to shoot me a note at singalminded@gmail.com.

That’s it! Again, if you’re paying for Singal-Minded you don’t have to do anything — your billing will pick back up when I relaunch the newsletter. New paid subscriptions will be closed in the meantime.

Thank you so much for reading. There’s a lot of good stuff to come.

Questions? Comments? Offers to preorder 15,000 copies of my book? I’m at singalminded@gmail.com.

New York Times health and wellness reporter Simar Bajaj has a new article about “The Return of Blaming and Shaming in Public Health,” as the headline puts it. It’s pegged to our esteemed Secretary of Health and Human Services, RFK Jr., and discusses the relationship between stigma and public-health concerns like obesity and smoking.

I don’t think there’s any good reason to be skeptical of Bajaj’s general thesis that stigma, in the sense of shaming or ridiculing or othering, is ineffective as a public health strategy. And there were things I liked about this article, which is a bit more nuanced than the headline suggests.

But I also think Bajaj makes a common mistake you see all the time in science writing, and it’s one that drives me a bit nuts. In arguing against approaches steeped in stigma, Bajaj writes the following:

Researchers found that antismoking messages focused on industry deception — like how Big Tobacco targets teens or manipulates the public — were some of the most effective at reducing cigarette consumption.

As soon as I read this sentence, I knew what would happen when I clicked on the link: I would be brought to a study that did not, in fact, meaningfully support the claim that we have such rich data about which antismoking messages work better than others.

How could I be so confident? Two reasons: First, we, as a species, don’t really know how to change people’s beliefs and behavior. Second, our hypotheses on this front are notoriously difficult to test scientifically.

Smoking itself is a fine example. The best way to know which anti-smoking messages work would be to take a huge group of otherwise similar people, divvy them up into groups that are otherwise as similar as possible, and randomly assign the different groups to be exposed to different messages about smoking. If, after the experiment ends, differences have emerged in the rates of smoking/quitting/other important behavioral variables between the groups, you might be able to infer that those differences were caused by the messaging.

But this is quite difficult to do in practice. It is difficult to find the resources to run a study this large and rigorous. And even if differences are observed, that’s not the end of the story: There could be all sorts of other explanations. The best-laid plans of researchers often go awry.

Of course the vast majority of studies aren’t nearly this ambitious. While we do have some randomized controlled trials of anti-smoking interventions, they are small and sometimes difficult to interpret. In this one, for example, 357 smokers in San Diego who had cigarettes delivered to their home by the researchers (not a bad study to participate in!) were randomly assigned to regular packs, blank packs, or to packs containing nasty imagery designed to deter smoking. The researchers found that “graphic warning labels decreased positive perceptions of cigarettes associated with branded cigarette packs but without clearly increasing health concerns. They also increased quitting cognitions but did not affect either cigarette cessation or consumption levels.”

One the one hand, a useful study, both because of the key result but also because it reveals something important and frequently ignored: People often profess beliefs that don’t translate into their actual behaviors. On the other hand — and this is not a knock on the researchers — even this fairly basic result is difficult to interpret! For example, “demand effects” (or characteristics) are a well-known phenomenon in experimental psychology — people often seem to catch on to the fact that they are being studied and to provide responses oriented (perhaps unconsciously) toward satisfying researchers. Let’s say I measure your beliefs about smoking, start sending you packs containing graphic warnings, and then measure your beliefs about smoking again. You report that you feel less positively about cigarettes. Was that because the imagery actually affected your perception of cigarettes, or because the experiment basically contains a giant blinding neon sign stating WE ARE TRYING TO MAKE YOU FEEL WORSE ABOUT CIGARETTES?

Also: There are huge cultural, political, and health-behavior differences between different parts of this great nation of ours. Would the results of a San Diego study, whatever they are, generalize to Peoria? Hell, would they generalize to San Francisco? It’s very difficult to know.

All of which brings me back to my skepticism of that New York Times sentence. If it’s difficult to establish reliably effective ways to change beliefs or behavior, and if in even the best cases (like an RCT) it’s difficult to interpret attempts to measure such interventions, it seems unlikely that anyone has actually established something as specific as “antismoking messages focused on industry deception. . . [are] some of the most effective at reducing cigarette consumption.”

Read more

May flew by! Here are some of your best comments, with some responses by me.

Jonathan Reale, on “On The New York Times’ White-Guy Discrimination Case (And Bubbles)”:

The rest of that quoted section from the Klein article also could benefit from some “turnabout” perspective. Imagine taking the attitude that because someone made a federal complaint, employees at the paper were justified in trying to ascertain and share the identity of the turncoat, with who knows what consequences to follow, if the racial category of the complainant had been something other than it is. I can imagine specific words/phrases that would likely get used to describe the group of employees and their activity as they vigorously sought to identify the traitor in their midst.

Yeah, I think there’s widespread skepticism that you’re “supposed to” use the federal civil rights law if you feel discriminated against as a white person. I have no doubt that this sort of discrimination has been vanishingly rare in American history as compared to other types but, as I pointed out in that article, it puts institutions in a legally tenuous position when they are basically advertising their desire to engage in it.

Snailprincess, on “It’s Immoral Not To Call The Cops On Genuinely Menacing People You Encounter In Public”:

I think a lot of this comes down to the fact that, in progressive circles, particularly online, it is in vogue to say “the system” is so broken it must be replaced. If that is your belief, then attempting to work within the system is pointless and morally wrong. But I’m going to say something controversial, the system is. . . mostly pretty good. Our system of courts, police, trials, courts, etc. is honestly pretty good, especially on paper. We absolutely fall short in execution plenty. Things can and should be improved. But no, we don’t need to rethink the entire concept of police. Almost all of the energy I see online is directed at “the system is broken, burn it down,” that there’s very little left for proposing actual reasonable improvements. I think that filters into politics as well. Very early on in the protests, there were some reasonably good proposals out of the “Black Lives Matters” protest, things like increased accountability and training for police, etc. Those got drowned out almost immediately by the “defund” rhetoric. Make America Incrementally Better Again.

I’ve been thinking about this in terms of some of the internecine fighting that goes on between (to oversimplify) the center-left and the far left. If you’re socialized to believe contemporary America is a world-historical hellhole that is mostly evil, then yeah, you’re going to want to burn it down. If — and this is closer to my perspective — you think America has some serious, frustrating problems, but in certain respects offers people a better life than almost anything that has come before it (here things get somewhat complicated with comparisons to Denmark and Germany, etc.), you’re going to be a bit more risk averse, especially when it comes to doing root-and-branch rather than piecemeal reforms of the justice system, “overthrowing capitalism” (whatever that means), and so on.

I’m probably a bit more bearish on our criminal justice system than Snailprincess, mostly because I think the idea of a two-tiered justice system is. . . basically true? That is, if you have money, you just have so many more opportunities to come out of scrapes with the system (civil or criminal) relatively unscathed than you do if you are lacking in resources. That is obviously unfair, though I don’t know how to fix it. I do think we’ve made tremendous progress toward the decarceration of nonviolent offenders and that we should acknowledge as such. Doesn’t mean you need to leap all the way to “never lock up any violent person for any reason unless you absolutely have to.” As I mentioned in the piece, sometimes it’s the only moral option to maximize everyone’s safety.

Justin DaMetz, also on “It’s Immoral Not To Call The Cops On Genuinely Menacing People You Encounter In Public”:

I recently had to call the police late at night here in suburban Tulsa because a homeless man had decided the space between my two cars in my driveway was the best place to sleep for the night. He was intermittently yelling and crying, and the noise woke my five-year-old. Being someone who is more progressive than not, part of me felt bad calling the police on someone who was clearly having a hard time in life already, and was just looking for a place to rest for the night. The last thing I wanted was to have some police officer having a bad night give him a tough time. But I was glad I did: One of the responding officers told me, along with being deeply inebriated, he also had a grazing gunshot wound that looked like it had been untreated. Calling the cops got him the medical care that might have saved his life from infection, and got him a better place to sleep for the night than my cold driveway. And the officers from TPD were all kind, calm, and compassionate the entire time they were at my house.

In some cases, contact with The System is better than the alternative!

J Chicago, on “How WPATH Thumbed The Scale Of Youth Gender Medicine Research”:

There is also a wider and serious nocebo effect. All these young people are being told they need these “life-saving” interventions in order to thrive, which was directly criticized after a UK study in response to the claims:

https://www.gov.uk/government/publications/review-of-suicides-and-gender-dysphoria-at-the-tavistock-and-portman-nhs-foundation-trust/review-of-suicides-and-gender-dysphoria-at-the-tavistock-and-portman-nhs-foundation-trust-independent-report

“The way that this issue has been discussed on social media has been insensitive, distressing, and dangerous, and goes against guidance on safe reporting of suicide.”

This is also discussed by Clayton:

https://link.springer.com/article/10.1007/s10508-022-02472-8

“However, an excessive focus on an exaggerated suicide risk narrative by clinicians and the media may create a damaging nocebo effect (e.g., a ‘self-fulfilling prophecy’ effect), whereby suicidality in these vulnerable youths may be further exacerbated (Biggs, 2022; Carmichael, 2017). This type of risk has been discussed in other similar situations involving youth (Abrutyn et al., 2020; Canetto et al., 2021; Shain & AAP COMMITTEE ON ADOLESCENCE, 2016).”

That is, patients who get these drugs might think they were saved from this risk.

And patients who don’t get them might be made even more anxious—a big problem right now in U.S. states where these interventions are not available for minors, and where the **MDs** treating them still believe that this means “lifesaving care” is being denied. The doctors pushing the “lifesaving” narrative are, in my opinion, doing harm to these kids by suggesting suicide is a natural response to having gender dysphoria and that gd can only be alleviated with these medications. It appears to me to be encouraging despair and fear, based on falsehoods.

Read more

Last week I published an article originally titled “How WPATH And The Endocrine Society Thumbed The Scale Of Youth Gender Medicine Research.” I say originally because I’ve since nixed “The Endocrine Society” from the headline. I’ve also unlocked that article, which was paywalled when I published it. What follows will make a lot more sense if you read that one first.

In the article I claimed that both WPATH and the Endocrine Society (ES) authored letters endorsing an observational design for the “Impact of Early Medical Treatment in Transgender Youth” study (those are links to PDFs of both letters), and that both letters exaggerated and oversimplified the state of quality evidence for youth gender medicine in 2014 (which, unfortunately, is not all that different from the state of quality evidence for youth gender medicine in 2026).

I made an error here: “The Endocrine Society letter was authored by the endocrinologist Wylie Hembree, who died in 2022 and who at the time was the Chair of the Society’s Task Force on Clinical Practice Guidelines for Transgender Care. He was also lead author of the 2009 guidelines. Writing on Endocrine Society letterhead. . . ”

I’m embarrassed I managed to screw this up given that I linked to the letter right in that paragraph, and it clearly isn’t on ES letterhead!

I’m honestly not sure what happened, but I apologize. I’ve corrected this in the text of the last article and, after this goes up, will include a link directing people here to explain what happened. (The WPATH letter clearly is on WPATH letterhead — believe me when I say I’ve now triple-checked.)

A couple people alerted me to this error. One was Dr. Michael Laidlaw, a California endocrinologist and youth gender medicine skeptic who has served as an expert witness for states seeking to ban youth gender medicine. He is also a co-author on the HHS umbrella review published last year, and the story of his attempts to get these documents is worth briefly telling.

In 2018 or so, Laidlaw developed concerns about the ethics of this research project. This led him to seek out information about it from the National Institutes of Health. He teamed up with the conservative legal activist Jane Robbins, a fellow skeptic, and she filed a Freedom of Information Act request for the Olson-Kennedy team’s NIH application packet that year. When it arrived they found it was missing two appendices they were interested in: the consent form that would be used to enroll children and youth in the study and — more importantly for our purposes — the two “letters of support” from WPATH and the Endocrine Society. Multiple follow-up FOIAs did not yield these documents. Eventually Robbins’ sister, Patricia Daugherty, filed a lawsuit against the NIH to obtain the documents, with the conservative legal group Judicial Watch serving as her counsel. “We reluctantly agreed to a dismissal since they at least produced the consent forms,” explained Laidlaw, but the other documents were still missing.

Laidlaw said he made a separate effort to get the letters via a California Public Records Act request to UCSF, a public university whose youth gender clinic is one of the four sites of the study (UCSF is also the home base of Olson-Kenendy’s co-investigator, the endocrinologist Stephen Rosenthal). That effort stalled out as well, and these dead ends were “what led me to call and email the Endo Society directly to try to obtain the letter that I thought they endorsed,” explained Laidlaw, who had been a member of that group since 2006.

I don’t remember how I first found out about the existence of these letters — it may well have been through Laidlaw. But for those of us who are curious about this project, they have long been an object of fascination. There really wasn’t any ambiguity about the letters’ provenance in the language of the grant application: Olson-Kennedy and her colleagues reference “the Endocrine Society and WPATH letters of support regarding our cohort design” and “Letters of support regarding our cohort design from both the Endocrine Society and WPATH.” Anyone reading the application would interpret this as the ES and WPATH lending their institutional imprimatur to Olson-Kennedy and her colleagues’ preferred study design.

That’s why Laidlaw, curious for the same reasons I was, reached out directly to the Endocrine Society in 2020. In October of that year, an ES staffer named Rodneikka Scott told Laidlaw the group had had no luck tracking down any such letter. “We double-checked with our colleagues in our Government and Public Affairs and others in an attempt to find the letter referenced in your email,” she told Laidlaw in an email that he shared with me. “As it turns out, there is no record of an advocacy-related letter of this same title.”

A few days later Laidlaw responded, gently insisting that the letter likely did exist given the clear references made to it in Olson-Kennedy and her colleagues’ grant application. “It seems to me to be a big problem if researchers claim they have a letter from the Endocrine Society that does not exist,” he also noted.

Scott then responded again, closing out the conversation:

Many thanks for your follow-up. Intrigued, I continued to dig a bit more into your request earlier today. The Endocrine Society does not write letters in support of specific grants. With that being said, we were able to find the attached two documents. The author of the grant may have attached one of these letters. We don’t know, though. I hope this is helpful as it is all that we have. [emphasis mine]

Neither of the two letters attached had anything to do with this research project, so those weren’t the ones in question. I have no idea why I got lucky where Laidlaw and Robbins struck out — why, late last year, in a long-overdue response to my follow-up FOIA request, the NIH sent me the two letters.

After I heard from Laidlaw, I emailed the ES earlier this week, quoting some of the group’s correspondence with him and pointing out that this shouldn’t be a complicated thing to establish: Either Olson-Kennedy and her colleagues’ claim that their letter of support had been authored by the ES was true, or it was false. In either case, I argued, the ES should clarify things.

I haven’t heard back. Just like I didn’t hear back when, prior to my publication of last week’s article, I reached out to the Endocrine Society to ask if the group wanted to comment or provide context about what was going on.

What we’re left with is a confusing situation. Johanna Olson-Kennedy and her colleagues repeatedly referenced a letter from the Endocrine Society in their NIH application. Such a letter would, of course, carry more weight than a letter from an individual endocrinologist — albeit an influential one — who merely had a role in the Society’s development of its gender dysphoria guidelines. The letter itself, however, looks much more like the latter: like the late Dr. Wylie Hembree was writing on his own behalf, noting his affiliation with the ES but not claiming to represent the opinion of that group on this particular matter.

I can imagine a critic accusing me of nitpicking, but shouldn’t researchers seeking millions in federal funding provide accurate information about which professional organizations have or haven’t written them letters of support? Shouldn’t the Endocrine Society, having been aware of this confusing situation since at least 2020 (via its correspondence with Michael Laidlaw), provide a public explanation? More broadly, should it have taken years worth of FOIA requests and a lawsuit just to obtain important documents that are part of a controversial, high-profile, and taxpayer-funded research project?

The most common theme in my reporting on youth gender medicine is a lack of transparency and accountability. It’s extremely rare for the researchers and activists whose work has come under scrutiny in recent years to be willing to answer even basic follow-up questions about their claims, their methodologies, or their theories. In some cases I can understand why — there have been lawsuits and harassment campaigns and all sorts of over-the-top ugliness. This is a very heated dispute. But when it comes to the top-tier researchers and research journals and professional organizations, I remain confused as to why they can’t answer simple questions about what they have done and why they have done it. These individuals and institutions often present themselves as the voices of scientific (and, frequently, moral) authority — they issue authoritative-sounding statements, they write op-eds, they get paid handsome sums to prepare expert declarations, and they consistently seek to influence ongoing public debates. It seems like they’re trying to have it both ways: They want attention and influence and prestige, but they don’t want to deal with the accountability that should come with those rewards.

I don’t want to let myself off the hook here for what was a very silly error. This shouldn’t have required two separate newsletters to explain, and I made the situation even more confusing than it needed to be. I should have noticed this apparent discrepancy and noted it in my first article. But fundamentally, the root cause of this confusion is a federal grant application that appears to contain false information. Someone directly involved should simply explain what’s going on here, but I won’t hold my breath waiting for that to happen.

Questions? Comments? More information about this? I’m at singalminded@gmail.com.

Update, 5/29/2026: It turns out that there’s some real uncertainty over the role of the Endocrine Society here. See here for a more information on that.

In 2014, Johanna Olson-Kennedy, medical director of the Center for Transyouth Health and Development at Children’s Hospital Los Angeles, was taking the lead on an ambitious research proposal known as “The Impact of Early Medical Treatment in Transgender Youth.” Olson-Kennedy and three other established youth gender-medicine practitioners — her co-investigators Stephen Rosenthal of UCSF Benioff Children’s Hospital, Norman Spack of Boston Children’s Hospital, and Robert Garofalo of Lurie Children’s Hospital in Chicago — were seeking millions in funding from the National Institutes of Health to conduct a yearslong effort to study the outcomes of youth who came through their four clinics. They planned to publish separate studies on the effects of puberty blockers and cross-sex hormones.

Suffice it to say that this sort of effort requires a mountain of paperwork, and I obtained some of that paperwork via a Freedom of Information Act request. The packet reveals that during the application process, the team hit a potential snag when the NIH reviewers evaluating the proposal expressed qualms about its being an observational rather than an experimental study.

This distinction is quite important for understanding what happened next and why it matters.

An observational study simply tracks a group of patients over time as they receive treatment. The problem is, if the patients’ conditions improve it’s difficult to know whether the medication itself is responsible. People who are sick often get better over time, without intervention. Or it could be that, while the subjects were receiving the treatment, they were also receiving other forms of care or advice from their doctors. There can also be strange and not fully understood mind-body mechanisms at work, including the placebo effect — sometimes people appear to get better simply because they think they’re getting treated. Observational studies might, in some cases, also be susceptible to the Hawthorne effect, in which patients who know they are being observed modify their responses to certain queries as a result. These problems can be addressed by attempting to statistically control for unobserved factors that might be contributing to patient outcomes, but there are real limits to what observational studies can reveal, and they can’t generally provide strong causal evidence about the efficacy of medical treatments.

An experimental study, on the other hand, often involves randomly assigning study participants to either receive the treatment being evaluated or something else. Sometimes the “something else” is a placebo — that is, half of the headache sufferers get a headache pill, while half get a sugar pill — and sometimes it is some other treatment that the new treatment is being compared to.

In the strongest, most idealized version of this type of study, every effort is made to treat the patients identically with the exception of the treatment in question, and patients, doctors, and almost everyone else directly involved are “blinded” as to who is in which group until the experiment itself is complete and it is time to analyze the results. In theory, this prevents patients in one group or another from being treated differently by doctors and others who interact with them. Researchers go to significant lengths to complete what are known as “double-blind randomized controlled trials,” the gold standard for medical research. In some cases, they even perform “sham” surgeries for the sake of RCTs.

In a letter responding to the NIH’s critique, Olson-Kennedy and her colleagues acknowledged that “by our estimation [it] was the most significant issue raised by the initial review,” but then argued against the idea of withholding puberty blockers from any of their study participants:

However, the current standards of care from the Endocrine Society and the World Professional Association of Transgender Health (WPATH) were developed from a panel of scientific experts and a consensus-building process based largely upon clinical experience and existing minimal research, and each supports the positive impact of intervention and the negative impact of no intervention. In the context of these guidelines, we strongly believe that a non-intervention control group, or certainly any control group as part of a traditional randomized trial, would be both impractical and unethical (see the Endocrine Society and WPATH letters of support regarding our cohort design in Appendix A). [Italics mine.]

It’s true that you can’t blind a study participant (or their family members) as to whether or not they have really received puberty blockers — not for long, at least. But methodologists have come up with ways to run solid, nonblinded (or “open -label”) experimental studies, so the choice isn’t just between a blind experimental study and an observational one.

Olson-Kennedy’s more important argument here is that, because the available evidence and guidelines “support[ed] the positive impact of intervention and the negative impact of no intervention,” it would be unethical to deny study participants access to puberty blockers.

It’s true that by the time Olson-Kennedy and her team were seeking funding, the Endocrine Society (2009) and WPATH (2012) had both published guidelines supporting the use of puberty blockers to treat gender dysphoria in at least some instances. But those guidelines — written by teams dominated by early practitioners of youth gender medicine — were not carefully written, as an independent Cass Review assessment later found (see online supplemental table 3). They did not include many of the guardrails that seek to attenuate the role of human bias in guideline development, and they recommended puberty blockers for gender dysphoric youth in the absence of quality evidence supporting this approach. Jennifer Block’s reporting in The British Medical Journal has more information about all of this.

The first group of clinicians to prescribe puberty blockers as a treatment for gender dysphoria certainly thought this treatment worked, despite that lack of evidence, and sought to codify that view into the relevant professional guidelines. Olson-Kennedy’s co-investigator Norman Spack, for example, was one of the earliest American youth gender clinicians to embrace puberty blockers, and he also co-authored the Endocrine Society guidelines. Given the prominence of the Endocrine Society, those guidelines went a long way toward legitimizing a medical protocol that, in 2009, could only be described as experimental. “There was an attitudinal shift to be able to say that the Endocrine Society supports this,” Spack told Endocrine News in 2013 (I came across that quote via the sociologist and youth gender medicine critic Michael Biggs).

There are good, well-understood reasons not to automatically trust physicians who believe a new treatment works but who lack data. That’s why the “evidence pyramid” underpinning the practice of evidence-based medicine puts “background information/expert opinion” at the bottom.

Doctors, like everyone else, are susceptible to bias. And absent a carefully controlled experimental setting, they don’t generally collect rigorous data on their patients, nor attempt to follow up with those whom they are no longer in contact with. All of which can give physicians a skewed understanding of the effectiveness of the care they are providing.

So while Olson-Kennedy’s argument to the NIH is larded with signifiers of authority — phrases like “panel of scientific experts” and the names of leading professional organizations — if you actually examine the timeline and the names involved, it really just boils down to “The small group of early practitioners of youth gender medicine believed puberty blockers worked as a treatment for youth gender dysphoria and enshrined their belief in the early guidelines.”

Yes, a couple of early, influential articles were published by the Dutch clinicians who pioneered the puberty-blocking protocol in 2011 and 2014, but they were methodologically weak observational studies of a carefully screened group of Dutch youth, and they did not tell a conclusive story: In many cases, the youth entered the study with solid mental health and they still had solid mental health when they were last observed. Even setting aside other, potentially backbreaking issues with this Dutch research that wouldn’t be recognized until years later, the studies didn’t come close to providing the quality of evidence required to justify the stance that non-intervention would be unethical.

That’s especially true given that, as Olson-Kennedy and her team explained in their application, they were seeking, in some cases, to begin puberty blockers at an earlier age than the usual Dutch threshold of 12 years old. As they wrote, “no data have been collected in the U.S. on the physiologic and mental health impact, safety, or tolerability of pubertal-blocking medical interventions… for transgender youth, particularly in children younger than age 12, leaving a gap in the evidence for this practice,” emphasis theirs.

Olson-Kennedy and her colleagues, then, simultaneously argued that there existed “no data” on the “mental health impact, safety, or tolerability” of blockers, particularly in this cohort — again, these are all their exact words — but also that it would be unethical to withhold this medication from any patients. These two claims are very much in tension.

As the team understood, 2014 was a crucial time to develop genuine evidence about youth gender medicine. And there were options available for an experimental study, at least hypothetically. For example, the Dutch clinicians had, before blockers came on the scene, provided counseling and other mental-health support to gender dysphoric teenagers until they reached age 16, at which point they could begin hormones if they were deemed to be good candidates for medical intervention (the counseling did not include attempts to convince the youth they weren’t trans, or anything like that). Olson-Kennedy and her colleagues could have at least attempted to run an experiment comparing that approach (or a version of it) to the newer protocol involving earlier medical intervention via puberty blockers. This sort of comparative study would not have been perfect, but it could have provided more useful data than an observational study.

Instead, they argued that an experimental study that withheld puberty blockers would be unethical, and to buttress their argument they pointed the NIH to “letters of support” from the same two groups that had published the early guidelines. The letters, which were not included in the original packet the NIH sent me but which a follow-up FOIA request yielded many (many, many) months later, are telling. They show that by 2014, despite a lack of quality evidence to support puberty blockers as a treatment for youth gender dysphoria, both organizations had decided not just that puberty blockers worked, but that they worked so well that it would be unethical to deny them to study participants. Given how much uncertainty remains about these treatments more than a decade later, this is a striking example of how individuals and institutions have, in their embrace of youth gender medicine, contravened normal standards of medical evidence.

The Endocrine Society letter was authored by the endocrinologist Wylie Hembree, who died in 2022 and who at the time was the Chair of the Society’s Task Force on Clinical Practice Guidelines for Transgender Care. He was also lead author of the 2009 guidelines. Writing on Endocrine Society letterhead, In his letter, he argued to the NIH that “Non-intervention in adolescents with Gender Dysphoria has been demonstrated to result in harmful physical and psychological outcomes, especially at the onset and during pubertal development,” and that “Including a randomized non-treatment arm would be unethical based upon existing data documenting adverse effects of failure to treat.” But whether non-treated adolescents with GD sometimes suffer bad outcomes is a different question from whether puberty blockers (followed by cross-sex hormones) effectively treat GD in the long run and are worth the potential downsides, and it was those latter questions (among others) that Olson-Kennedy and her team were seeking to answer. (Correction, 5/26/2026: The letter wasn’t written on Endocrine Society letterhead, but Hembree did list himself as “Chair, Endocrine Society Task Force on Clinical Practice Guideline Transgender Care,” and Olson-Kennedy referred to the letter as having come from the Endocrine Society. I’m reaching out to the Endocrine Society again to attempt to clarify this.)

The WPATH letter was written by Jamison Green, then the president of WPATH and the chair of its Ethics Committee (as well as a big name in trans activism more generally), and by Jamie Feldman, chair of the organization’s medicine and research committee. Both had been involved in the development of the 2012 WPATH guidelines. The letter read, in part, “There is no question that pubertal blockers block puberty,” the implication being that an experimental study would be unnecessary. The authors also argued that “We have generations of transgender people who have been denied puberty blocking treatment, and we know the trajectory of these people’s lives.” Citing the high suicide risk of transgender youth, they argued, “The prospect of entering a program for treatment with a chance that treatment will be withheld could endanger the mental and physical health of the subjects.” This is question-begging, given that Olson-Kennedy and her team were specifically investigating whether puberty blockers would ameliorate suicidality among study participants. (I reached out to both the Endocrine Society and WPATH for comment. I received a call from a public relations firm representing WPATH and, after a quick off-the-record chat about this story, didn’t hear anything further. I didn’t hear back from the Endocrine Society.)

There are, of course, situations in medical research where withholding treatment would be unethical. But that’s only the case when there’s a great deal of certainty about the treatment in question (the relevant technical term here is equipoise, or a situation of genuine uncertainty about which approach is best). There was no such certainty when it came to the use of puberty blockers: There were always obvious potential side effects worth grappling with concerning infertility, sexual dysfunction, bone health, and cognitive development, as critics of youth gender medicine have noted. Olson-Kennedy and her team, as we saw earlier, were well aware of these and other uncertainties, and even the aforementioned WPATH guidelines point out that “the long-term effects [of blockers used for gender dysphoria] can only be determined when the earliest treated patients reach the appropriate age.” It’s simply a misunderstanding of basic bioethical principles to claim, under conditions of such profound uncertainty, that it would be unethical not to treat.

These letters demonstrate that the Endocrine Society and WPATH both endorsed the idea that puberty blockers were already an established, effective treatment for gender dysphoric youth — that many of the very questions Olson-Kennedy and her colleagues were seeking to answer had already been answered. (Which, of course, would raise the question of why they needed federal research dollars in the first place.) They were not convincing responses to the issues raised by the NIH. Nonetheless, the NIH allowed the study to proceed with an observational design — no comparison group.

In August 2015, Olson-Kennedy and her co-investigators received an initial $5.7 million to begin their project. More than a decade and about $10 million later, they haven’t come close to answering the key questions they set out to answer. Their most important publication, a 2023 New England Journal of Medicine study on hormones, is rife with cherry-picking, outcome-switching, and other signs of shoddy and/or politicized science. And even if you accept the team’s highly suspect methodological choices, their results do not offer compelling evidence to support the administration of hormones to adolescents with gender dysphoria — not least because there is no comparison group, and the authors did not control for therapy and/or medications like antidepressants.

As for the blockers study, in 2024 Olson-Kennedy told Azeen Ghorayshi of The New York Times that “Puberty blockers did not lead to mental health improvements… most likely because the children were already doing well when the study began,” as paraphrased by Ghorayshi. Olson-Kennedy admitted that she and her team had withheld this data out of fear that it would be weaponized by conservatives seeking to ban youth gender medicine. The final version of that study still hasn’t been published, but a preprint of it (that is, a pre-peer-review version) published online last year appeared poised to repeat many of the mistakes of the 2023 NEJM study.

The failure of gatekeepers here can’t be overstated. WPATH and the Endocrine Society have attempted to present themselves as rigorous-minded groups that follow the science, but in this instance they acted much more like activist organizations (not for the last time, especially in WPATH’s case). Of course, it didn’t help matters that the project’s lead, Olson-Kennedy, had already publicly described these treatments as lifesaving years before she began studying them: As she told ABC News in 2011, “We often ask parents, ‘Would you rather have a dead son than a live daughter?’ ”

At the risk of repeating myself, I can’t emphasize enough how foolish it is to blindly “trust the experts” unless the experts in question have earned that trust. There are myriad ways the difficult process of scientific inquiry can go off the rails, and “The Impact of Early Medical Treatment in Transgender Youth” is a prime example.

Questions? Comments? I’m at singalminded@gmail.com. Image via the Endocrine Society.

I feel bad for the anonymous 23-year-old woman interviewed by the New York Post earlier this week. She explained that on April 2, she and her friend were attacked and followed on the subway by a menacing man who wouldn’t leave them alone: “the crazed suspect stalked them closely and allegedly yanked her by the back of her head in an attempt to slam her to the ground and booted her friend in the back.” Thankfully, during their escape they came across some NYPD officers and the suspect was immediately arrested.

The woman and her friend eventually declined to press charges, she told the paper. “Maybe a part of me was just like, I don’t want to put another black man in jail,” she said. That’s the quote that’s gone viral since then: I don’t want to put another black man in jail. It’s gone viral because a month later the black man in question, Rhamell Burke, pushed retired schoolteacher Ross Falzone down a flight of stairs in Chelsea, killing him.

Reporting from The New York Times and the Post paints an astonishing story. Burke had, under the name Rhamell Burke-Missouri, been a Broadway dancer before descending into profound mental illness during Covid. The timeline of last Thursday almost defies belief: On the same day, he was released from a psychiatric unit, killed Falzone, and showed up for a scheduled court appearance on another assault charge. (Though he had been identified by then, no one realized he was wanted for a new crime during the court appearance, according to the Post.)

It’s a cropped version of that quote from the anonymous Post source that’s circulating. In full, it reads, “Maybe a part of me was just like, I don’t want to put another black man in jail, but, you know, at some point, if you are a criminal, you’re a criminal, and he was scary, he was a scary guy.” The shorter version of the quote is being slightly misrepresented online, as though it proves that if she and her friend had only intervened then and there, the later killing never would have happened. But as the last line of the story notes, “Burke was hit with an assault charge for the April attack and granted supervised release at arraignment.” I’m guessing, based on her quote, that the anonymous victim feels that if she had cooperated more, his punishment would have been more severe, averting the killing of Falzone. That’s far from certain, and I don’t entirely blame her for her decision. Even if her quote indicates one particular motivation, there are all sorts of reasons to not want to cooperate fully with a prosecution.

I do think, however, that we should pause and think through which of these two options — help the police or decline to do so — is more morally fraught in a situation involving a clearly dangerous man (it’s almost always a man) threatening people in public. If you do help the police, you of course have no control over where the perp will end up, what sentence he will receive, and so on. There’s real moral uncertainty because you’re not saying “I choose to help the police and in doing so I hereby declare that the proper punishment is 90 days of inpatient treatment at a gleaming, freshly constructed facility whose rec room has nice river views.” That isn’t how it works. Neither option is close to perfect.

The incident involving this Post lady reminded me of the wave of online outrage sparked by Emma Vigeland, co-host of The Majority Report, in 2023, over comments she made on her show about the killing of Jordan Neely by Daniel Penny — another instance in which the failure to keep a dangerous, deranged man off the streets led to an unnecessary death (his, in this case):

I was hit at one point sitting on the subway by a man who was having a mental health episode. He sat next to me, and he was elbowing and kind of flailing around. He hit me in the face and in my body. It was jarring, right? The idea that I would want him to be hurt in any way — I just didn’t want to be near him in that moment because I understood something was going on here. Every one of us who has taken public transit has had this kind of situation, something similar happen — seeing someone struggling. That doesn’t mean that our fear in that instance — and I was a little scared, because I was hit —but my fear is not the primary object of what we should be focusing on right now. It’s the fact that this person is in pain. And so, like the politics of dehumanization, privileges, the bourgeois kind of concern of people’s immediate discomfort in this narrow, narrow instance, as opposed to larger humanity and life, it’s really frickin’ twisted.

Now, I’m not an impartial arbiter here because I dislike Vigeland, and no online pile on delivers the “correct” amount of scorn, but of course it’s ridiculous to treat “not wanting to get assaulted in public” as a “bourgeois kind of concern of people’s immediate discomfort.”

Vigeland’s comments here reflect a worldview that is staunchly opposed to police involvement: You should hardly ever call the police on anyone because the police are very likely to make things worse. This norm has been taken to dizzying extremes during moments of peak concern about police abuse. At one point during the summer of 2020, for example, the head of a small Twin Cities literary agency tweeted that she had called the police in response to a gas station near her getting looted. This sparked a minor uproar online that led to three of her staffers publicly resigning, because (apparently) it’s racist to call the police on looters.

But let’s stick to situations involving genuinely menacing people on the subway. Not people who seem a little off, not people slumped against a wall within a station, but people who wander around, in public, threatening others or actually hurting them, as in the case of Vigeland and the Post lady. These types are fairly rare, at least in the relatively safe parts of New York City where I spend the most time, but in my experience they have become less so since the coronavirus. (The difference between “safe-” and “unsafe”-feeling neighborhoods is also blurred a bit by the extent of our subway and bus coverage.)

My argument is that if you are assaulted or seriously menaced by someone like this and you don’t get the police involved, you’re doing something wrong — in fact, you may be wronging not only this person’s future victims but the person himself.

Read more

The U.S. Equal Employment Opportunity Commission is suing The New York Times on behalf of a veteran journalist who claims he was denied a promotion due to being a white male. Both the lawsuit itself and the response to it are revealing.

New York Magazine’s “LOL Get a Load of This Loser” Approach Is Tired and Unhelpful

A useful entry point is New York magazine’s short write-up, in Intelligencer, by features writer and media columnist Charlotte Klein. The web headline that pops up on Google reads “The White Man Crying Discrimination at the New York Times” (the live headline on the story itself is milder).

Klein’s story does not devote any real space to helping readers understand whether and to what extent the suit has merit. Rather, the assumption seems to be that everyone reading the article is on the same page: This is ridiculous.

This makes a certain level of superficial sense given our political moment. Donald Trump really likes to weaponize the courts against his perceived enemies. He has used seemingly every conceivable legal lever, in seemingly every conceivable manner, to attempt to make life difficult for institutions and individuals he feels have wronged him. He has done this both in his personal capacity, launching plainly baseless multibillion-dollar defamation lawsuits against outlets like The Wall Street Journal (dismissed for now) and the Times itself. Perhaps more menacingly, he has sought criminal prosecutions against political enemies like New York attorney general Letitia James (repeatedly) and former FBI director James Comey (for posting a photo of seashells* spelling out “86 47” in seashells on a beach to Instagram, which contra Trump’s DoJ, is clearly not a violent threat unless you think “86 the scrambled eggs,” when overheard at a diner, is a waitress’s call to murder scrambled eggs). While these cases tend to falter one way or another because they are (to use the technical legal term) ridiculous, of course that doesn’t matter to the individuals and institutions targeted, who are facing a vindictive, erratic world leader with effectively unlimited resources at his disposal. (Correction: This paragraph originally said Comey himself spelled the offending numbers out in seashells. This is false: He says he and his wife found them already arranged like that. His is being charged for posting a now-deleted photo of the arrangement to Instagram. Thanks to the commenter who alerted me to this error.)

Trump and his deputies are also particularly outspoken about their desire to use the cudgels of federal authority to wage culture wars, most notably against the universities that they view as having been taken over by DEI lunacy. The same general trajectory applies to the EEOC, which, as Klein notes, is “now controlled by a Trump appointee who has vowed to help wage the president’s war against DEI culture[.]”

If I were coming to this controversy fresh, I’d probably assume this EEOC lawsuit was politically motivated and had no merit, simply because the Trump administration is basically a giant factory that produces meritless, politically motivated legal actions. And even despite what I’m about to say, I can’t even deny the possibility of a political motive here! You don’t think Donald Trump or anyone seeking to ingratiate themselves to him would lick their chops at a chance to sue the Times? The EEOC has significant discretion to choose which of the many complaints it receives to investigate and potentially elevate to full-blown lawsuits, and wherever there’s discretion, there’s the potential for politicization and other forms of bias.

The thing is, I’m not coming to this controversy fresh, and neither is Klein. She knows, as do I, as does anyone who worked in a progressive institution during the peak of the circa-2020 racial reckoning, that during this period there was a large amount of open, legally precarious talk endorsing employment discrimination against white people — particularly white men. Everyone in our world knows that this occurred, and in some cases, is still occurring. I cannot bold this enough! If you are from our circles, you should pause and consider the merits of a case like this — not immediately slot it in alongside Trump’s (and Kash Patel’s) many abuses of the legal system. (This article by Jacob Savage generated a lot of discussion when it was published late last year and is a good starting point for anyone skeptical that what I’m describing actually occurred.)

Now, legally precarious is not a synonym for illegal. We’ll dive a bit more into the complaint in a moment. It’s here, it’s just 17 pages long, and you can and should read it. (Klein doesn’t link to it, and as Robby Soave points out, this is an unfortunate tendency among outlets writing about litigation). But I want to linger for a moment on Klein’s coverage. Because if you read her piece, you will not find yourself edified one iota as to the underlying merits of the disgruntled employee’s claims. Rather, you will be immediately buffeted by politically loaded signifiers, giving this whole story the stench of yet another instance of Trump and his toadies unfairly attacking the media. That is, Nothing to see here other than more MAGA grotesquerie.

Naturally, there are quotes from Times insiders supporting this view:

The paper itself was the first to break the news of the suit but did not name the employee who made the complaint. Reporters at the paper have been scrambling to figure out the employee’s identity, driven in part by bafflement that one of their own colleagues would sell out the paper to the administration, which has used tools of the federal government to attack the press. “You’re giving the Trump administration a weapon while they’re trying to persecute journalists,” said one reporter.

“This has been kind of a shitshow behind the scenes — people trying to figure out who the aggrieved person is,” said another Times staffer.

. . .

People at the paper say the claim is absurd. “I’m sorry, there are plenty of white guys at the top of the New York Times. Not really something that’s holding you back,” said the reporter. To name one prominent example, Joe Kahn, the paper’s executive editor, is a white male, as are many members of the masthead.

That last graf tells you all you need to know about Klein’s approach to this story. First, “People are saying/People say” is a classically suspect journalistic construction, not to mention a pretty funny one to use in this context. “People” say a lot of things! Especially nonrandom samples of them.

The quote which follows is, as Soave notes, a total non sequitur. This case centers on one staffer/applicant’s claim that he was discriminated against. At no point in the suit do the EEOC attorneys claim that the Times doesn’t have white men in positions of senior leadership. This blatantly fails the turnabout test: Klein would never cover a lawsuit claiming anti-black discrimination at a company by quoting a skeptic pointing to the existence of black higher-ups at the company in question. Because Klein does not take this complaint seriously, entirely different standards of evidence and reasoning apply.

I don’t doubt that there are employees at the Times furious that this employee filed this complaint. They are the ones most likely to speak to Klein about it. But I do doubt that that’s the whole story. For years, there’s been an ongoing conflict within the Times — sometimes rather heated — between staffers who take a more swashbuckling approach to matters of social justice and staffers whose views on the proper roles of newspapers and journalism are a bit more traditional, and it would be surprising if we didn’t see the same cleavage here.

I’m positive that, especially among staffers who think that social justice activism within the Times became unreasonable and interfered with the paper’s mission (a sizable group that was able to assert itself a bit after the reckoning had crested, as this earlier coverage by Klein makes clear), the reaction to this lawsuit is significantly more complex than anger at Yet Another Aggrieved White Man. I would bet significant money that there’s a cohort of them rolling their eyes not at the complaint, but at the sloppiness of the Times’ approach to DEI — that they view this more as an inevitable instance of the chickens coming home to roost than as an employee acting in a reactionary or unreasonable manner.

The Bubble

Back in 2021 I wrote about epistemic bubbles, about how in certain communities members are socialized to believe that certain arguments are compelling. Outside those communities? Less so. So one of the downsides of being inside an epistemic bubble is that you come to believe things that — whether or not they are true — are unconvincing to outsiders. (See also my more recent column in The Dispatch.)

In communities centered on social justice, the inspiring, unifying sense of being “on the right side of history” can sometimes overshadow more terrestrial concerns. And if you disagree with the community’s preferred approach to some issue on the grounds of those more terrestrial concerns, that calls into question whether you are, in fact, on the right side of history in the first place, which in turn calls into question the legitimacy of your group membership. This can, in the most extreme cases, bring reputational or even professional consequences.

Within progressive institutions, the racial reckoning was quite intense. I would argue that — as is often the case during periods that feel revolutionary — the decisions institutions made about how best to meet the moment were not particularly well thought out. At the time, it was actually quite difficult to openly disagree with whatever the activist-minded members of your organization were saying, simply because it was so unpleasant to find oneself on the wrong side of that revolutionary fervor.

The reckoning brought with it accelerated demand for certain preexisting diversity policies. One such demand centered on diversity trainings, which, as I argued in the Times in 2023, might do more harm than good and be legally problematic, especially in their more intense, psychologically demanding circa-2020 incarnations (indeed, these trainings did spark some legal complaints). But it was very hard to say so as interest in these interventions was peaking! Revolutionary fervor and all that.

Another long-standing demand, newly reenergized during the reckoning, was to diversify workplaces — the subject of the EEOC complaint. The question of which sorts of interventions to diversify a workplace are legal is complicated. It centers largely on Title VII of the Civil Rights Act, which contains a straightforward proscription on hiring, failing to hire, promoting, or failing to promote someone on the basis of their “race, color, religion, sex, or national origin.” Over the years the courts have carved out some exceptions, but they’re quite narrow. So if a complainant can prove that race was a factor in their suffering a negative employment outcome, in theory they could have a strong Title VII case. Race doesn’t even need to be the only factor.

Of course, countless institutions have tried to become more diverse while staying within the boundaries of this law, and it’s the courts that make the final kosher/haram determination. This has never been a simple matter but it has become more of a moving target at the moment because of a couple of Supreme Court cases: First, there’s 2023’s Students for Fair Admissions, Inc. v. President and Fellows of Harvard College, which effectively ended race-based affirmative action in college admissions and which may come to have significant ramifications in employment settings as well. More straightforwardly, 2025’s Ames v. Ohio Department of Youth Services (that one decided 9–0) rejected a “background circumstances” rule that made it harder for members of majority groups to bring successful Title VII challenges. That rule assumed that discrimination against majority groups in employment settings was “unusual” and imposed a higher burden on members of these groups seeking to argue that they had been discriminated against. Absent the obstacle posed by that rule, it became significantly easier to claim anti-white discrimination in employment settings.

All that being said, a newspaper like the Times does have options at its disposal to increase workplace diversity that are probably legal if certain precautions are taken. A newspaper is allowed to beef up its recruitment at historically black colleges and universities, to invest heavily in developing pipelines that extend all the way down to (say) majority-minority high schools, and so on. Efforts to expand the pie of potentially qualified applicants have been generally blessed by the courts as legal under civil rights law, though there are always potential landmines: If you send a rep to an HBCU career fair as part of your diversity efforts, for example, but that rep refuses to let a white student write down their email address to be notified about job openings, that could be legally risky, because it could be seen as a proscribed form of discrimination on the basis of race.

Much more legally problematic are policies like rigid quotas, which clearly entail discrimination in decision-making for a given position. (A super-simplified example: A company has two manager-level positions and has decided one of them should be black. If there’s an opening for one position at a time when the other is held by a white employee, then, definitionally, anyone applying for the open position who isn’t black will be discriminated against on the basis of their race.) And of course you, the employer, face grave legal risk if you enact or express preferences — in either direction — about the racial identity of the successful applicant for an advertised position. Should a complaint arise, that could be seen as explicit evidence of discriminatory intent.

In many cases, the question of whether a given intervention is on safe legal ground has to do with a fiendishly complex array of legal, regulatory, and political factors. The NFL, for example, instituted the so-called “Rooney Rule,” in which NFL teams are required to interview candidates for certain high-level coaching and management positions who are external to the organization and who are female and/or members of minority groups. This creeps right to the line of at least a certain sort of quota, albeit a quota involving a given applicant pool, not final employment decisions. In theory, this is legal under Title VII — probably? But you don’t know for sure until you test it in the courts, and testing it in the courts usually requires a plaintiff with “standing” (that is, someone who has grounds to sue). No one has challenged the Rooney Rule on the basis that it discriminates against white applicants, which doesn’t, on its own, really tell us much about its legality. It would be difficult for a rejected white applicant (the only sort of person with standing to bring such a case) to prove they had been discriminated against, and it would be career suicide to do so given that the NFL is a tiny and hypercompetitive labor market. (That said, Florida’s conservative attorney general recently threatened to sue the league over the rule.) All of which is to say that the absence of lawsuits does not mean a given diversity policy is on legally sturdy ground.

There are clearly a ton of complexities here, especially given the Trump administration’s aggressive approach to waging the aforementioned battles and, longer-term, a conservative-dominated Supreme Court that has expressed a severe allergy to race-based decision-making in academic and employment settings. The present case doesn’t really involve these complexities, though.

Read more

It was a hectic month! I’m about to fly across the country again! There’s a lot going on and several deadlines collided last week, so you won’t see responses from me this time around. I apologize for that. You guys had a lot of good and though…

Read more

I’m late to this, but when the host of Jeopardy! name-checks you, how can you not write about it?

It was a negative name-check, to be clear. In March, Jennings wrote on Bluesky, “Jesse Singal: ‘I don’t understand why all these experts with degrees keep disagreeing with me. So demoralizing. What could the explanation be??’ ”

I need to engage in an annoying bit of pre-explaining (presplaining), because I deleted the tweet that precipitated all of this. I’ll relegate a fuller explanation to a footnote, but the short version is I did not mean to imply I favored the Ken Paxton policy in question. Rather, my post was a response to Jack Turban’s claim that there is a clear scientific consensus on the subject of youth gender transition — a claim he was not making for the first time.1

Jennings, of course, got famous as a wildly successful contestant on, and then as the host of, a game show where almost every question has a single correct answer. (Well, it’s Jeopardy!, so technically every answer has a single correct question.)

Unfortunately, a sizable subset of the progressive world, in my experience, believes that extremely complex scientific disputes are more or less like Jeopardy! What’s the answer to a question? If there’s any ambiguity you consult a panel of judges — The Experts. Whatever The Experts say is the truth of the matter, and you can win an argument by citing the existence of an expert, or experts, who agree with whatever claim you’re promoting.

Does youth gender medicine work? Yes. What’s the detransition rate? Extremely low — like, wildly so. Are teenagers’ views of their gender identity influenced by social and cultural factors? Of course not. How do we know all of this? Because “the leading experts” agree. And you can, of course, find experts who will make these claims. In Jennings’ view, if you disagree with the experts, that is prima facie evidence something is wrong with your position, because experts are generally not to be questioned.

This conflation of “expert view” and “correct view” is common in progressive circles. The experts are basically the oracles: You go to them for the truth of the matter, as spoken directly by the science gods. There was a certain brilliance to the trans activist talking point that “the experts all agree” that youth gender medicine works, because respect for and deference to mainstream experts is burned into the modern progressive soul. Overwhelmingly, we’re the ones who both populate the nation’s sensemaking institutions and consume their output: We’re far more likely, compared to conservatives, to become university professors and journalists and public-health officials and to read scientific papers and pop-science treatments of controversial disputes. It’s one of the key ways we differentiate ourselves, whether the subject at hand is evolution, anthropogenic climate change, or youth gender medicine. We trust the experts; they do not.

To be sure, “trust the experts” works in some situations. We trust engineers to build bridges since the vast majority of the time, bridges don’t collapse. The development of the smallpox vaccine really did allow us to completely eliminate smallpox from the wild. I trust a trained engineer over some random guy to build a bridge, and I would not trust anyone who claimed the smallpox vaccine didn’t do what the experts claim it does. More generally, certain scientific debates do basically have two sides, with one being right and the other wrong. The small numbers of people who continue to believe in young-earth creationism or to not believe that humans cause climate change are wrong about the facts of the matter.

But these sorts of right-and-wrong science disputes generally center on very well-understood natural processes, very well-researched diseases and vaccines, and so on. Most matters of genuine, ongoing public scientific controversy are significantly more complex than that. Sometimes the science itself is unclear, sometimes the attendant questions of values and trade-offs are fiendishly difficult, and sometimes both. (There is no “right” or “wrong” answer to how long schools should have remained closed after the peak of the pandemic, for example.)

The contemporary progressive view of science, as reflected by Jennings, sweeps a huge amount of genuine complexity and controversy under the rug. Its adherents vastly overestimate the number of genuinely settled scientific controversies as well as the ability of scientific findings alone to resolve complex policy disputes in the first place. In doing so, they give “experts” far too much credit, and they give the average educated person far too little credit for their ability to challenge experts. On the whole, this understanding of science is profoundly elitist and disconnected from the messy realities of how research is practiced, and who is treated as an “expert” in the first place.

It’s Not Only Okay to Question Experts — It’s Necessary for a Healthy Public Science

Let’s take Jack Turban as an example, since it was his name that sparked this whole thing. Turban is impeccably credentialed: He’s a psychiatrist and professor at the UCSF School of Medicine, he has degrees from Harvard and Yale, he did his residency at Stanford, and he is regularly cited in major media outlets as an expert on youth gender medicine. As his bio page notes, he has won awards like the Excellence in Public Health Award from the U.S. Preventative Health Service, and he was named a Pride 30 Innovator by NBC.

In Jennings’ view, this is the platonic ideal of the trustworthy expert. Who would disagree with someone like that?

I think this is silly, elitist (I’m going to keep using that word because that’s really what this comes down to, more than anything), and wrongheaded. I think the average educated person can be trained to understand enough about the underlying concepts to either judge for themselves whether Jack Turban is trustworthy, or to be able to do a reasonable job interpreting public criticisms of Turban. (In some fields, that wouldn’t be the case. I don’t think the average educated person can understand complex disputes about quantum mechanics. That’s not what this is, though.)

Of course, it would be nice if this were unnecessary. It would be nice if, during a given dispute between multiple parties, we could apply some sort of mathematical formula that adds up everyone’s degrees, weighs for prestigiousness, and tells us who to trust. If that were the case, surely Jack Turban would win by sheer dint of the number of illustrious institutions he is affiliated with. But this would be very unwise.

To see why, let’s start with an extreme example: Just Some Blogger. If there’s anyone you shouldn’t trust on matters of science, in the Jennings/Bluesky world, it’s Just Some Blogger, right? Just Some Blogger is their bête noire, and for often solid reasons: When it comes to issues of genuine scientific misunderstanding, like the debunked link between vaccines and autism, who has done more damage than Just Some Blogger? Worse still are Just Some Blogger’s descendants, Just Some YouTuber and (ew) Just Some Substacker.

Except: You can’t always trust the impeccably credentialed expert over Just Some Blogger!

In 2021, as Turban was becoming an increasingly well-known name among those familiar with the youth gender medicine controversy, someone named JL Cederblom published a post to Medium titled “The Lukewarm Perjury of Jack Turban,” which is a deep dive into a single sworn statement Turban published as an expert witness in one of the legal cases surrounding a youth gender medicine ban. JL Cederblom is classic JSB — the name is a pseudonym, and we, the reader, have no idea what his or her credentials are. Why would you possibly trust “JL Cederblom,” Just Some Blogger of unknown provenance, over Jack Turban, MD, he of Yale and Harvard and Stanford?

Because, if you read their post, you will see that JL Cederblom is clearly in the right here: They are correct that Turban routinely bungles and/or misrepresents fairly basic facts in this particular document. I’m not saying I agree with every single claim in this long, bloggy takedown, but on the whole, you really can read Turban’s claims, Cederblom’s responses, the source material (usually studies from the literature), and recognize that Turban’s sworn declaration puts a thumb — maybe several thumbs — on the scale.

Or you can do what I did a while ago: Listen to Turban’s interview on Science Vs where he touted the strong evidence base for these studies, track down the studies he and the show referenced, and read what they actually say. If you do, you’ll find that Turban is, again, presenting an exceptionally rosy view of the literature.

Or — and I’m giving you multiple options here, to demonstrate that none of this requires a PhD in statistics — you can look up other sworn statements Turban has made on this subject and see how they hold up. For example, much of this debate comes down to the question of study quality, to the differences between strong studies and weak ones. This is an area known as evidence-based medicine, or EBM.

Leor Sapir, a Manhattan Institute critic of youth gender medicine who led the and the lead author on the Health and Human Services report about it, published a piece in 2023 arguing that Turban appears unfamiliar with basic EBM terms, including an absolutely vital one: systematic reviews. I can imagine Jennings saying: “What! Outrageous!” Sapir has a PhD, sure, but in political science. He isn’t a medical doctor like Jack Turban is and he definitely lacks Turban’s esteemed pedigree and history of expertise on this subject. Who the hell is he to claim that the good doctor lacks such basic knowledge? (Correction: Sapir emailed me to say he isn’t considered the lead author of the project. It’s a subtle but meaningful-enough-to-correct difference.)

But again, we have a choice here: We can do what Jennings implied as the correct course of action in his post about me — defer to whomever seems like the “expert” — or we can try to actually evaluate the disagreement and see who is correct. I have read the deposition in question, and Turban, answering questions posted by a legal adversary, under oath, really doesn’t seem to understand what a systematic review is.

On the one hand, it’s shocking, given Turban’s lofty perch as a Leading Expert who makes hundreds of dollars an hour producing sworn declarations and sitting for sworn depositions on this subject, that he doesn’t really know what a systematic review is (or didn’t, at least, at the time of the deposition in question). On the other hand, to be honest, I wasn’t that surprised, because it’s been clear to me that Turban shoots from the hip in a sloppy way since at least 2019. In November of that year, Turban and his frequent co-author Alex Keroughlian published a Psychology Today article that claimed “As recently as 2013, the Diagnostic and Statistical Manual of Mental Disorders [DSM] referred to being transgender as a ‘gender identity disorder.’ ” This is a completely false claim, albeit (for very in-the-weeds reasons) a frequently made one. The day after his piece went up, I pointed out to him on Twitter that there was no textual support for how he was describing this debate (I think my quote here is from his now-deleted tweet):

It only took Turban and Keroughlian and/or Psychology Today a year, but they eventually corrected the false claim from their piece:

It’s fairly wild that a public-intellectual psychiatrist, who should be more acquainted with the DSM than anyone else, would make such an obviously wrong argument about it. But before Turban and Keroughlian relented and made this correction, you could have played the same elitist game Jennings wants you to play: Why should you trust some sleazy, just-asking-questions journalist over two clinician-researchers as stuffed to the gills with Credibility and Expertise as Turban and Keroughlian? Maybe because sometimes people are treated as experts despite not exhibiting the competence of experts? Because sometimes people are elevated to the status of “expert” because they make politically expedient claims that fit certain cultural grooves, not for reasons of pure merit?2 Because we don’t have a formal state priesthood?

Again: Jack Turban is not a quantum physicist wowing his audience with astonishing strings of mathematical notation. He is a doctor making pretty straightforward, pretty specific claims about what various documents and studies do or don’t say. Some of these documents and studies require some training to be able to understand, but for the most part this debate is not, contra what Ken Jennings would have you believe, so far out of reach of the average educated reader that we have no choice but to mindlessly defer to experts.

Maybe on its own we could just call the DSM blunder a random slip-up on Turban’s part, the sort of brain fart we all experience from time to time. But you combine that with his under-oath misunderstanding about systematic reviews, and the issues highlighted by JL Cederblom, and so much other stuff that I don’t even have the time to get into here. . . I think Ken Jennings is way off base in instructing me — and, more importantly, his huge, fawning audience — to automatically view it as suspect if an “expert” like Jack Turban disagrees with me.

These Guys Don’t Actually Trust Experts Anyway

In a sense I’m giving Jennings and his Bluesky audience too much credit, because they don’t actually trust experts in the sense of automatically deferring to them on the basis of their credentials. Rather, Jennings and his audience first decide what the “correct” view is on a given subject, then they search out the experts who support that view, then they pretend those are the only “experts” in the room.

Jennings has only fleetingly chimed in on youth gender medicine, but he clearly views any skepticism of it as beyond the pale. In fact, after California governor Gavin Newsom said in an interview he was torn on youth gender medicine — not that he was against it, but that he was torn — Jennings agreed with a trans activist who said she would never support Newsom because of that stance. “Any candidate cynically ‘triangulating’ on trans kids is a non-starter,” wrote Jennings, “and now is the time to say so. There’s still so much time to advance candidates that DON’T suck.”

Jennings has decided, then, that any skepticism about youth gender medicine constitutes immoral “triangulating” on what he clearly views as a civil rights of such exigence that it rises to single-issue-voting status for him: allowing minors access to puberty blockers, hormones, and double mastectomies.

Jenning did not adopt these beliefs because he reviewed the scientific evidence and found it so compelling. Rather, he adopted this belief for the same reason most of us adopt our beliefs — matters of tribal affiliation, peer pressure, and so on — and then he reverse-engineered his way into his belief that “the science is settled” or whatever. Of course, there are now many experts who feel differently, who have decided that the science is very unsettled, because the systematic reviews that have been conducted tell such a clear story to that effect. In my experience, youth gender medicine advocates reach into a bottomless sack of excuses and smears to explain away these experts — anyone who disagrees is a bigot, made howlingly obvious scientific errors, and on and on. In the worst cases, massively credentialed, super-respected “experts” have simply lied — there’s no other word to describe what they did — to denigrate their perceived opponents. Jack Turban himself, for example, put his name on this. That’s who Ken Jennings wants me to defer to, like some sort of robot — a guy who will sign his name to lies without a trace of discernible shame.

So in one sense this entire exercise is bad faith — Jennings only trusts the experts only to the extent the experts reflect his own beliefs back at him. But since he’s such a famous and well-respected guy, I wish he wouldn’t promote such silly, superficial, and elitist ideas about the relationship between the average citizen and scientific “experts.”

Questions? Comments? Online petitions to get me on Jeopardy! so I can hash this out face-to-face with Jennings during what would be the most unwatchable TV segment ever broadcast? I’m at singalminded@gmail.com. Image: BEVERLY HILLS, CALIFORNIA - MAY 02: Ken Jennings attends UCLA Jonsson Cancer Center Foundation's 28th Annual Taste For A Cure at Beverly Wilshire, A Four Seasons Hotel on May 02, 2025 in Beverly Hills, California. (Photo by Robin L Marshall/Getty Images)

As soon as the novel Upward Bound came out a few weeks ago, two things happened: It became a runaway hit — now a New York Times bestseller — and a bunch of critics loudly wondered whether it was really written by its ostensible author, Woody Brown.

In case you haven’t heard about this controversy yet, Woody Brown is a 28-year-old man with severe autism. He suffers from echolalia, or the involuntary recitation of words or phrases, and his utterances are often drawn from the children’s programming he has a fixation with. He employs a “media cart” with multiple screens playing this sort of content to keep him calm.

Well, it’s not that he employs the media cart. That’s the thing: It’s his mother, Mary Brown, who is his caretaker. Brown communicates with her by pointing at a letter board using a technique called the Rapid Prompting Method (RPM).

This will all sound familiar if you’ve heard of facilitated communication, or FC. FC is a broader set of practices that purportedly allow severely disabled people to communicate via the use of keyboards or letter boards, usually with the help of a caretaker who holds their arm steady. This practice was propelled along by an undeniably seductive idea: that a bunch of people who had been written off as incapable of communication due to their disabilities in fact had rich inner lives that could be unlocked by the proper caretaker wielding the proper technique. There was a surge of media attention, a surge in demand for FC training, and a sudden uptick in the number of severely disabled nonverbal people who were now communicating — sometimes in remarkably sophisticated ways — thanks to FC.

Unfortunately, it wasn’t real. Debunked isn’t a strong enough word for FC — disgraced might be a better fit. The technique appears, in many cases, to be driven by a Ouija board type of effect in which it is actually the facilitator who produces the words attributed to the nonverbal person in question. At the peak of FC, some of the words that were produced were very consequential. There were false sexual allegations. There was an unbelievable scandal involving a “consensual” relationship between a professor and disability activist named Anna Stubblefield and a nonverbal young man. Rigorous trials have shown, rather conclusively, that FC does not work as advertised.

Woody and Mary’s method, RPM, is basically just FC. It has not been rigorously evaluated. When I reached out to Soma Mukhopadhyay, the technique’s founder — and the woman Mary Brown credits with unlocking Woody’s apparently enormous potential — she sent me a few papers that didn’t come close to demonstrating what needs demonstrating: that RPM generates text that originated with the disabled person, not their facilitator. (I mentioned this in a premium episode of Blocked and Reported we produced on this controversy.)

RPM has taken Mary and Woody on a remarkable ride. His communicative potential unlocked — and with Mom at his side facilitating his communication — he allegedly got an undergraduate degree at UCLA, he allegedly followed it up with an MFA in creative writing at Columbia, and he allegedly wrote Upward Bound.

I say “allegedly” because no one knows if he actually accomplished any of these things. That’s what’s so remarkable about this story — how long it’s been going on with so little pushback from anyone who could have, well, pushed back. The Woody Brown controversy didn’t emerge ex nihilo. Rather, there’s a lot of history and context here, and that history and context should cause any reasonable observer to default to the explanation that it’s Mary, not Woody, doing the communicating here. (It doesn’t help matters that you can watch videos of Woody “spelling” that rather clearly demonstrate he is not, in fact, spelling the words Mary is attributing to him, as Jill Bearup and others have demonstrated.)

Daniel Engber, who wrote the definitive Stubblefield story for The New York Times Magazine all the way back in 2015, has done more investigation into Upward Bound than any other journalist. In his Atlantic article, he mentioned “the phalanx of established authors who have mentored Brown and endorsed his work”:

Those who responded to my questions told me that they’d found no reason to suspect that he had not written what they’d read. Rivka Galchen, a staff writer at The New Yorker and an associate professor at Columbia, worked closely with both Brown and his mother across four semesters [during his MFA program]. Although it had crossed her mind, at first, that his writing might be influenced, the worry vanished over time, based on what she saw. “I’m not a doofus,” she told me. And even if some doubts had lingered, she would have felt both unqualified and disinclined to investigate the question. “Do I have students whose girlfriends write their prose? Do I have students who use AI? I have no idea,” Galchen said. “I feel like I have to take it on faith.”

One more:

Mona Simpson taught Brown in two classes at UCLA and supervised his senior project, an early draft of Upward Bound. “He has a natural instinct for the shape of a story,” Simpson told me. “I truly have no doubts about Woody’s authorship.” But over the course of our conversation, she acknowledged the vagaries of collaboration — the possibility of some interpretation at play: “It could be that they’ve worked together so long that she can intuit some of what he’s intending. I don’t know.”

These responses are glib given what’s at stake here, but I can also understand how this puts individual instructors in a tough decision. Asking tough questions about Woody and Mary Brown is clearly the right thing to do, but it feels like the wrong thing to do. Wouldn’t it be much easier to assume that their communication system works, that someone, somewhere, had done the required rigorous testing? It’s a form of very human buck-passing, especially because it isn’t individual instructors or professors, after all, who make decisions about disability accommodations — it’s administrators.

I reached out to UCLA and Columbia University about this. Now, these institutions have certain obligations under federal law regarding Woody’s privacy. I wouldn’t expect them to renege on these obligations. Instead, while disclosing that I was writing about the Woody Brown controversy for this newsletter, I simply asked what their more general policy was toward facilitated communication. I sent both schools’ media offices versions of this email:

Read more

Boston magazine just published an article about Steven Pinker headlined “Can Steven Pinker Save Harvard?” Subheadline: “But the celebrity professor’s own record raises a question: Is he the right guy for the job?”

First of all, I don’t get that “but.” It’s not referencing anything! It’s like the original headline was going to be something like “Steven Pinker Wants to Save Harvard,” and then someone changed the headline without changing the subheadline.

Setting aside my overreaction to a minor copy-editing error, this conceit is also a bit much — it’s very magazine-y. No one, including Steven Pinker, thinks Steven Pinker is (single-handedly) going to “save Harvard.” The article is really about a few different things, most of them summed up in the very first paragraph: “His critics call him a cover for racists,” writes author Robert Huber. “He’s been accused of providing intellectual ammunition to the alt-right, and of dismissing inconvenient evidence when it doesn’t fit his theories.”

Huber doesn’t dive deeply into the idea that Pinker has “dismiss[ed] inconvenient evidence” — he mostly just quotes a couple of academics who have made that claim. So I’m more interested in the (re-re-re-re-)litigation of the idea of Pinker providing cover for or legitimizing bad actors. There’s a very long, involved controversy here that mostly has to do with virulent anger, on the part of a small but loud group of academics, about Pinker’s status as one of the leading public-intellectual debunkers of “blank slatism,” or the idea that all humans are basically the same, and that any differences that emerge are the result of external factors.

I find it surprising, in 2026, that adherents of the more sweeping anti-Pinker view have done so poor a job of addressing counterarguments to their position (I’m going to table the narrower and more standard academic debate over whether he has gotten this or that wrong in his books; obviously, it’s legitimate to closely read and critically respond to the work of as influential a figure as Pinker). Their myopia on this matter can, I think, be explained by their own form of blank slatism. They believe that people are more or less blank slates, with regard to political opinions, until they decide which scientific beliefs to adopt. Similarly, political ideologies are only adopted because they are seen as having scientific legitimacy.

So, the argument goes: Without figures like Pinker, who are at best useful idiots and at worst quiet but intentional enablers, the alt-right would have far less intellectual fuel and wouldn’t have gained the power it has gained. Or if they aren’t arguing this, I don’t understand how they could possibly have remained so mad at Pinker for so many years.

But there’s another way of looking at things: Humans have certain hardwired or nearly hardwired tendencies. We’re very good at carving the world into us versus them, for example. Is there a society anywhere that doesn’t do this? There are a million strains of us-versus-them-ism. There are relatively “soft” versions like civic nationalism, in which differences that in other contexts might matter a great deal are overlooked (or partially overlooked) in favor of a big collective identity. That’s America, at its best, though there are obviously also downsides when it comes to xenophobia in times of war, and so on. But civil nationalism is cupcakes and cotton candy compared to the harsher, more evil forms of us-versus-them-ism — racism, Nazism, and so on.

Read more

Below the paywall you’ll find my latest column in The Dispatch. This is one of the (many! uncountable!) benefits of being a premium subscriber to Singal-Minded — you get these columns about a day after they appear (also behind a paywall — sorry everything in the world is paywalled) on The Dispatch.

This column is a response to a conspiracy theory that bothers me. The conspiracy theory holds that the backlash to trans rights is the result of “out-of-state billionaires” (Graham Platner’s view) or center-left/“reactionary centrist” types, myself included, in outlets like The Atlantic or The New York Times (all of Bluesky’s view).

The basic idea is that voters have been bamboozled into thinking a “fake” issue is a “real” issue. There isn’t actually anything to talk about when it comes to the briefly, wildly successful progressive campaign to change how society views sex and gender identity, and to institute many of those changes into law and regulation (and despite the backlash, many of those codified changes remain).

I’d like to briefly mention the response to the article from Alejandra Caraballo, the trans activist and clinical instructor at Harvard Law School’s Cyberlaw Clinic. She has come up before in this newsletter and is generally not worth engaging with. It is to our collective benefit that she has sequestered herself on Bluesky, where she can do her thing all day far from society.

But I can’t resist responding to two bits from the lengthy skeetstorm she posted in response to my article (you won’t be able to read it unless you’re signed in — which, maybe for the best…).

First:

Singal frames trans advocacy as demanding “someone’s sex is what they say it is” with “no questions asked.” This is the caricature that the right spent those hundreds of millions constructing. He’s repeating the propaganda in a moderate voice to appeal to liberal’s prejudice. [next skeet]: Actual legal frameworks for gender recognition vary enormously and involve documentation, clinical engagement, and administrative processes. Singal collapses all of that into the most inflammatory possible framing and attributes it to a monolithic movement. It’s a strawman.

Specifically, I wrote that the trans backlash appears to be the result not of some broad, deep-seated animus toward trans people and a desire to see them removed from their homes or jobs (for example), but because the “specific arguments trans advocates and progressive politicians have chosen to embrace simply don’t appear to have worked at all,” emphasis in the original.

And that, I argue, is

largely because they center on a maximalist version of self-ID, or the notion that someone’s sex is what they say it is. Blue states have passed laws extending this logic to areas like locker rooms and prisons, and these are wildly unpopular positions for obvious reasons: Most people don’t think biological males should be able to enter female spaces with no questions asked.

It is undeniably — undeniably — the case that mainstream trans activism has pushed for no-questions-asked self-ID! I don’t understand how anyone could have the audacity to claim otherwise! It’s too bad the word gaslighting has been ruined by overuse because damn, is this close: An activist coalition pushes X, someone responds by noting that X is very unpopular and leads to some bad outcome, and one of the activists in question sputters “What? We weren’t pushing for X! What are you talking about! That’s the most inflammatory possible framing!”

Obviously, Caraballo is correct in saying that the “actual legal frameworks” vary, but that’s not due to a lack of effort on the part of activists, who have pushed for just about the simplest legal framework possible: Because this person says so. California itself is a great example: It used to have a system that provided prisons with the discretion to make individual decisions about where trans women should be housed. Then the ACLU and other activist groups pushed for a law that basically turns this into a matter of self-ID, with explicit “regardless of anatomy. . . based on the individual’s preference” language that makes it difficult to keep males out of women’s prisons.

This from Caraballo annoyed me a lot more:

Singal claims marriage equality imposed “nothing other than hurt feelings” on opponents. This is revisionism and ignores religious exemption litigation like Masterpiece Cakeshop. Opponents made the exact same tangible-harm arguments. Singal sanitizes that history to make the comparison work.

Here’s the passage Caraballo is quoting from: “With a very tiny number of exceptions, gay marriage did not, in fact, impose anything other than hurt feelings and offense on its opponents.”

Part of the reason to stay away from Bluesky is that it is filled to the brim with the most brazen, shameless liars. I can’t imagine a better example than this: I write a sentence in which I make a point, note the existence of exceptions to that point, and link to the most famous exception. Caraballo responds by claiming I didn’t address the exception, quoting from the sentence where I addressed the exception! What do you even do with a “public intellectual” who acts in this manner? Again: Thank God for Bluesky and its sequestration function.

Subscribe now

Anyway, the column itself is below for premium subscribers. Hope you enjoy it.

Trans Issues Are No Conspiracy

The public pays attention to trans issues because trans activists wanted them to.

There’s been an undeniable backlash against the trans rights movement. In February, the liberal outlet The Argument reported on the results of a survey it had commissioned of 3,000 registered voters. The problem wasn’t just that the mainstream trans advocacy position was underwater on genuinely fraught issues like sports and youth gender medicine, but that even a seemingly settled issue—bathrooms—also now polled poorly for trans people.

Read more

If the “controversy” over Joe Rogan endorsing Bernie Sanders hadn’t actually happened, it would sound like something a right-wing internet troll had fabricated. It was that idiotic.

This was back in 2020, when, as you recall, professional progressives were hard at work at their nationwide Make As Many Americans As Possible Hate Us initiative — an undeniable success. On a then-recent episode of his podcast, Rogan had expressed enthusiasm about Sanders. “I think I’ll probably vote for Bernie,” he said, as quoted by CNN. “Him as a human being, when I was hanging out with him, I believe in him, I like him, I like him a lot.”

Rogan, then and now one of the most successful podcasters in the country, not only has a huge audience, but a huge audience of the very group Democrats have struggled with in recent cycles: young men. Naturally, then, the Sanders campaign tweeted out the clip of Rogan endorsing their man.

This, as hard as it might be to believe, sparked a “controversy.” Rogan, you see, had opinions that did not line up 100% with progressive orthodoxy, including the most important, vital-to-defend sort of orthodoxy: orthodoxy that had just been established. CNN dutifully reported, for example, that Rogan had

questioned — using offensive language — whether a transgender MMA fighter should be able to compete against other women. [This is a reference to Fallon Fox, a male-to-female MMA fighter.]

“If you want to be a woman in the bedroom and, you know, you want to play house and all of that other sh-t and you feel like you have, your body is really a woman’s body trapped inside a man’s frame and so you got a operation, that’s all good in the hood,” Rogan said. “But you can’t fight chicks.”

The Human Rights Campaign put out a statement denouncing Sanders for accepting the endorsement, noting that “In 2019, 25 transgender and gender non-confirming [sic] people were killed because of the type of transphobia that Rogan stokes.” One relevant fact about HRC here is that the organization vigilantly searches out stories of trans people who have been murdered, retcons these individual tragedies as having been caused by transphobia even when there’s no evidence that was the case, and then uses these misrepresented killings as a fundraising device. (I wish I were making this up.) “Given Rogan’s comments, it is disappointing that the Sanders campaign has accepted and promoted the endorsement,” tutted moral exemplar Alphonso David, who was president of HRC until whoops. “The Sanders campaign must reconsider this endorsement and the decision to publicize the views of someone who has consistently attacked and dehumanized marginalized people.”

If you take what HRC and a bunch of shrieking online weirdos — some of them, then as now, in positions of institutional progressive power — were saying at face value, it would mean that candidates like Sanders are allowed to accept endorsements only from figures whose views are already quite far to the left. This would, of course, make it very difficult to win any sort of national election.

On his podcast, Bill Simmons sometimes dips into his “Conspiracy Bill” persona, which grants him license to float wildly silly conspiracy theories about, like, the NBA draft lottery. Allow me here to briefly 1) don my Conspiracy Jesse hat, which is actually a velvet fedora; 2) note that if you’re living a comfortably upper-middle-class life working at a D.C. organization like HRC, some tiny part of you must know that your professional life actually benefits from Democratic losses; 3) ask you not to get offended because this is a simple corollary to the obvious fact that, for example, anti-abortion groups were undeniably hurt by Dobbs; and 4) take off the fedora and emphasize that I don’t really think this.

Anyway: We’re seeing a different version of this same controversy now, and it concerns a figure who is not seen as too far to the right but as too far to the left: Hasan Piker. The background here is slightly complicated but it comes down to the fact that, as per Politico, the centrist group Third Way is putting pressure on Democrats to not associate with Piker after Politico revealed that Piker “will stump in Michigan with Senate candidate Abdul El-Sayed next month.” Other big lefty Democrats, like Bernie Sanders and Alexandria Ocasio-Cortez, have previously done events with Piker, but now the pressure appears to be ramping up:

“Piker is close to — but not over — the Nick Fuentes line, where going on his show itself is indefensible,” Third Way co-founder Matt Bennett, who’s been sounding the alarm about Democrats’ affiliation with Piker, told POLITICO. Bennett added that Democrats “take on all of his baggage if they don’t overtly reject” him, which he said is “dangerous because it empowers the right and gives them an incredibly powerful tool to hit Democrats with that’s very bad.”

This strikes me as extremely naive. The idea that if you appear on a stream with someone, or do an event with them, you “take on all [their] baggage” is the sort of thing you could only come to believe if you are an overpaid political analyst. I just don’t think it accurately captures the relationship between influencers and their audiences. A lot of people tune in to folks like Piker and his ilk not because they agree with everything they say, but because of the parasocial feeling of hanging out with someone whose general vibe or worldview they like. This is difficult for those of us who are over a certain age or who don’t consume a lot of streaming (or both) to understand, but I strongly believe it’s true. This is just an extension of the commonly voiced but rarely internalized-among-political-elites finding that the “average American” doesn’t have the same sorts of meticulously labeled ideological commitments as college-educated, politically activated types. Again, it’s more vibes: We’re the weird ones for giving ourselves specific political labels and for holding bundles of beliefs that don’t actually “go” together for any coherent reason.

Read more

There are some significant changes coming to Singal-Minded. I’m excited to go into more detail soon, but one bit of very low-hanging fruit I have thus far failed to pluck is the excellent comments you guys post under each and every one of my articles. That changes now: Going forward, I’m going to publish a monthly roundup. In some cases I’ll simply highlight comments that deserve to be highlighted, and in others I’ll also respond. Feel free to use the comments sections of these posts to continue the conversation further, of course.

Thank you to [NEW MYSTERY ASSISTANT EDITOR WHO I WILL MORE FORMALLY INTRODUCE IN THE NEAR FUTURE, OR NMAEWIWMFIITNF], for doing the bulk of the work of compiling these.

Randall Randall, on “The Horseshoe Always Wins: Gettin’ Mad At Headlines Edition,” has thoughts on a not-great tweet from CNN I mentioned in that piece.

The CNN tweet was remarkable in that it seemed to assign responsibility for the “drastic change” to the arrest, rather than to the throwing of bombs. It doesn’t even say they threw bombs, just that they were arrested for it! This is the kind of wording many have come to see as an attempt by a media outlet to imply that someone did X, when the article then buries near the end that they were only accused of doing X. It’s language that signals something misleading is being conveyed, and it’s up to the reader to figure out how and what the writer’s bias was.

Ann, also on “The Horseshoe Always Wins: Gettin’ Mad At Headlines Edition”:

Genuinely confused by this statement: “I’m not calling for outdated ideals of perfect objectivity or whatever —” Perfection is not possible, agreed, but is it not still considered a worthy goal?’

All I was getting at here was that I don’t expect journalists to be robots viewing situations from 50,000 feet. It’s silly to demand we never let any moral judgment seep into our writings, but it’s even sillier to demand we inject large amounts of moral judgment — especially when the heinousness of an act or an attempted act speaks for itself.

Andrew Wurzer, on “AI Is Too Consequential For Half-Baked Lefty Critiques”:

AI is a great force multiplier. It’s a terrible replacement — so far, anyway. Can it replace entry-level employees? Some of them, at least. Enough of their tasks that they may not be worth hiring in nearly such large numbers.

The difference between an entry-level person and an LLM is that the entry-level person can reliably learn to get much better and more independent in a few years. The LLM? Jury’s still out. I have my doubts, but it’s certainly not out of the realm of possibility. But if we stop developing higher-level talent, and AI is not able to keep up, in ten years we’re going to have problems as experienced people retire and cannot be replaced because we didn’t train anyone to do it.

The problem with all of this is that it’s very hard to keep up with the progression of LLMs. In my experience, LLMs can get better over time — and/or, perhaps just as importantly, their human users can get better at knowing how to usefully prompt them, when not to trust them, and so on. And the improvement does seem to be pretty insane over time. It’s just hard to say exactly which walls in which domains they’ll hit. In my daily use of Claude and ChatGPT, I’m consistently blown away both by how good it is and the basic nature of its errors! Just today I asked it a question about a specific date range I circled in red on a screenshot of a calendar. It confidently generated a long, likely compute-intensive answer based on totally different dates, and then I had to correct it! I just think so much money is on the line that those sorts of basic errors are likely to become less frequent in time.

I’d like to think that firms and organizations won’t be so myopic that they’ll fail to understand which tasks can’t be outsourced to AI and leave themselves in the situation Andrew describes. It’s going to depend on the humans responsible for making these decisions and their ability to accurately assess what can and can’t be outsourced. The general name of the game still seems to be expert human supervision for anything mission-critical that’s more complicated than entry-level.

TK-421, on “A Response To Critics Of My AI Article And An Apology To Librarians”:

There’s an old Edsger Dijkstra quote: “The question of whether a computer can think is no more interesting than the question of whether a submarine can swim.” I wouldn’t put it quite that strongly — I think it’s a fascinating philosophical question in the right context — but in terms of evaluating whether a computer can perform a particular task or have a particular impact on society, whether it’s “really thinking” or not is much more of an academic question.

I’m agreeing with TK-421 agreeing with me, but this is well-put. It’s interesting to think about the Chinese room thought experiment, then versus now. The basic tension always centered on the fact that the guy inside the system clearly didn’t know Chinese — he was just blindly following instructions. Even though the common response to this is that the entire system the guy is immersed in “knows” or “understands” Chinese, it strains our intuitions a bit because the whole thing seems so clunky. But when the thing doing the translating/conversing/writing/reading is lightning fast and its inner workings are obscured to us, the so-called functionalist arguments just feel different.

Read more

A huge number of public debates are fundamentally pointless and, at least arguably, scammy.

What I mean is this: People express extremely strong opinions about certain subjects without ever explaining exactly what they mean. They’ll say they are ardently pro-X or anti-Y. But if you ask them to explain exactly what they mean by X or Y, you’ll often be met with hostility, the request itself treated as prima facie bad faith.

This sort of half-baked conversation benefits a lot of demagogues. These sorts of people are good at coming up with creative and often vitriolic ways of expressing their ardent support for X or their opposition to Y and, just as importantly, their disdain for people who feel differently. But if they were forced to carefully and transparently define their terms, they’d have a lot less control over the ensuing conversation and might find that they are not intellectually or rhetorically equipped to engage meaningfully in it.

Take the (related) debates over Black Lives Matter and DEI (diversity, equity, and inclusion) programs. When these terms first went viral, good-versus-bad dichotomies emerged. To say you were in favor of BLM or DEI was to say you endorsed progressive views on race, policing, and the like. Likewise the opposite: To oppose BLM or DEI was to express support for conservative views on race, policing, and the like. It was all rather performative.

Eventually, the debates got more complex — it became clear that these terms could be seen as “floating signifiers” that meant different things to different people. Clarification became important. Did endorsing BLM mean that when police shoot someone, they should be subjected to a rigorous and independent investigation? Or did BLM mean, as the website BlackLivesMatter.com at one point endorsed, “disrupt[ing] the Western-prescribed nuclear family structure requirement by supporting each other as extended families and ‘villages’ that collectively care for one another, especially our children, to the degree that mothers, parents, and children are comfortable”?

I’ve found the DEI debate even sillier and even more in need of clarification. Do you “support DEI”? Shouldn’t it depend? If supporting DEI means ensuring everyone follows civil rights law, supported by trainings centered on the nature of those laws, that’s one thing. If supporting DEI means intense workplace encounter sessions in which employees are forced to cop to being (perhaps unconsciously) racist, perhaps exacerbating rather than ameliorating intergroup tensions in the process, that’s another thing. To this day, though, you see a lot of pundits, academics, and others pretend that being for or against DEI is an inherently good/bad position, as though this means much of anything in the absence of a lot more information. It’s a floating signifier and also a political football.

***

Earlier this month, the researcher David Manheim wrote on X, “People keep repeating ‘stochastic parrot’ — often without any mental process behind it to specify what the argument is.” So he decided to do something about it: He wrote “a paper dissecting various possible arguments, and explaining which are valid. Here’s a blog-post version.” He published that version to the rationalist website LessWrong. It’s titled “Hunting Undead Stochastic Parrots: Finding and Killing the Arguments.”

For those unfamiliar with the stochastic parrot argument, it comes from a very influential 2021 conference paper published by Emily M. Bender, Timnit Gebru, Angelina McMillan-Major, and Margaret Mitchell (going by “Shmargaret Shmitchell” here) titled “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?”

I promise this isn’t another Singal-Minded piece about AI. It’s a piece about how to think and write clearly. But, briefly, Bender and her colleagues wrote that “Contrary to how it may seem when we observe its output, [a language model] is a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot.”

Subscribe now

Even though the paper is about a lot of other stuff as well, because stochastic parrot is both in the title and is an ingeniously memorable phrase, it has turned into its own sort of viral floating signifier. You’ll commonly see arguments of the form that AI can’t really think or reason or understand, because it’s just a stochastic parrot. In my experience — and to borrow slightly from my last post on this subject — these arguments are often very vague, depending entirely on how the arguer him or herself would define terms like think or reason or understand. It’s also gotten harder and harder to treat these models as mere parrots as they’ve become increasingly powerful, unless we’re talking about some sort of mutant super-parrot (again, to be fair, the article was written in 2021).

Manheim’s post starts:

I argue the “stochastic parrot” critique of LLMs is philosophically undead — refuted under some interpretations, still valid under others, and persistently confused because nobody defined it clearly. This is an attempt to fix that.

He then lays out seven different versions of the “stochastic parrots” claim, all of which have different meanings.

He continues:

The most practically significant observation from this taxonomy is that there’s a conflationary alliance, a term Andrew Critch coined, among groups skeptical of LLMs, built on the ambiguity of “stochastic parrots.”

People with materially different philosophical and empirical commitments can all say “LLMs are just stochastic parrots” and mean completely different things. The Markovian version appeals to critics who wish to claim LLMs are simple. The Social Normative version appeals to critics who think society needs accountability from the LLMs. The Teleological version appeals to those dismissive of increased agency, and to safety researchers worried about goal-directed systems. Clearly, the groups have overlapping rhetoric but widely divergent implications — they’d give different advice about what to do, what evidence would change their minds, and what would constitute a solution.

The stochastic parrot argument was always a cluster of distinct claims, some of which were valid about earlier systems and have since been refuted, some of which remain live but depend on specific philosophical commitments, and some of which are unfalsifiable. Once these are separated, most of the ammunition for the argument disappears — what remains are legitimate concerns about accountability and social norms (Social Normative SPs) and a contingent empirical debate about generalization robustness (Unreasoning and Optimization-Artifact SPs), rather than a fundamental barrier to LLM understanding.

It doesn’t matter if you, personally, don’t get all of this. Parts of this paper were a bit above my own head, technically speaking. The point is that this is an extremely useful and intellectually helpful way of approaching this conversation. It advances it in a way merely having the ten-thousandth yes/no argument about “stochastic parrots” doesn’t.

The next person who wants to come along and participate in this debate can look at Manheim’s list and do all sorts of. . . stuff. They can refute the claim that anyone’s even arguing for this or that version. They can argue that versions Manheim is claiming are dead or ill are actually alive. And so on. This post, and the paper it will soon turn into, is fundamentally productive. Whereas a lot of pundits and a distressing number of actual “intellectuals” use slipperiness as a substitute for actual reasoning and argumentation, Manheim is actually intellectual and has reputational skin in the game, making arguments that are rather specific.

This might sound basic, but I’ve been depressed by how much of intellectual life — again, including intellectual life as carried out by actual intellectuals — operates at a much lower, fuzzier, more demagogic level than this. People like Manheim should be lauded for engaging in productive intellectual work, even if it turns out to be wrong, and the scamsters whose reputations rest on slipperiness should be called out as such.

As public (pseudo)intellectual life gets increasingly fractured and balkanized and shot through with demagogues and scam artists, I’m becoming increasingly interested in finding signals of trustworthiness. In general, if someone makes a strong claim — especially a morally weighted one — and then refuses to define their terms or answer follow-up questions, I think this is an extremely useful heuristic: This probably isn’t someone worth taking seriously. Of course this, like any other heuristic, isn’t 100% accurate, but if you’re out here attempting to build your reputation and/or intellectual project by making public claims, what excuse do you have to not meaningfully elaborate on and defend them?

Questions? Comments? Follow-up questions and requests for clarification? I’m at singalminded@gmail.com. Image: 07 March 2026, Berlin: A pair of parrots play at their nesting place in an alder tree on the Landwehr Canal. The collared parakeets (Psittacula krameri) have been at home in a park in Kreuzberg for two years and are very popular with walkers. They are extremely robust and can therefore spend the winter in freedom in cities. Photo: Jens Kalaene/dpa (Photo by Jens Kalaene/picture alliance via Getty Images)

I was invited to write a chapter for the recently released book Viewpoint Diversity: What It Is, Why We Need It, and How to Get It. I was very glad to do so, because my thinking on this subject has changed over the years — in 2020 I wrote about “Why I’ve Changed My Mind About The Importance Of Ideological Diversity In Academia,” an article which briefly recounts why I didn’t used to take this issue seriously but why I now do.

I chose to write about the implicit association test, which I’ve reported on at length for New York magazine and in a chapter in my own book. In this new chapter, which builds off of points I’ve made previously, I theorize that the IAT may have spread as far and wide as it did due to a lack of ideological diversity (or viewpoint diversity) in social psychology. The IAT and its surrounding marketing told a story that left-of-center scholars were very sympathetic to: Unconscious racism is rampant and has a significant negative impact on society. In a field with more viewpoint diversity, I argue, this sort of thing may be less likely to occur. It’s an example of how viewpoint diversity can perhaps have a salutary effect on deeply troubled areas of science.

Inside Higher Ed just published a searingly negative review of the book by a writer named John K. Wilson, and my own chapter features prominently. Wilson comes at these issues from a very different perspective than I do: In 1995 he wrote a book called The Myth of Political Correctness, and a glance at his author page on IHE suggests he spends a lot of time fighting back against the idea that there are any major climate issues on campus with regard to speech (at least climate issues caused by those on the left). That’s completely fine. What’s less fine is the dishonest way he approached my chapter.

Here’s the entire excerpt in question from his review:

In his essay “Viewpoint Diversity Can Kill Zombie Ideas” (which is an attitude directly contrary to everything viewpoint diversity should stand for), Jesse Singal asks, “What conceivable downside could there be to introducing more political diversity to the process, some way or another? Could it possibly make things any worse?” The first rule any libertarian like Singal should know is that it’s always possible to make things worse. In particular, attempts by government officials and administrators to impose “diversity” in faculty hiring almost invariably make things worse, particularly when “some way or another” is the mechanism imposed.

Singal’s careless indifference to the potential for political repression of higher education would be awful at any time, but coming in the midst of the Trump administration’s actual political repression, it reveals his breathtaking indifference to censorship.

I’m really baffled by the claim that wanting to kill zombie ideas is “directly contrary to everything viewpoint diversity should stand for.” Zombie ideas flourish in situations where they cannot be openly challenged!

Anyway, it’s never a good sign for the quality of a review when you, the person being reviewed, learn new things about yourself from your reviewer. Just from these two paragraphs: I didn’t know I was a libertarian, I didn’t know the chapter in question had anything to do with faculty hiring, and I definitely didn’t know that I had exhibited “careless indifference to the potential for political repression of higher education,” though to be fair, I suppose that last one isn’t something anyone would openly admit to.

The main problem here is that Wilson gives no context whatsoever about the actual subject of my chapter. That’s how he can reprint the quote “Could it possibly make things any worse?” and choose his own definitions for it and things. So I’m simply going to publish the full context from which Wilson is quoting here, and then readers can decide if he is faithfully rendering my case. This is not exactly a thundering-from-the-pulpit chapter. By the excerpt in question, I’ve already written that “I want to make a speculative case that viewpoint diversity within academia could serve as a rather reliable inoculation against these occurrences” and “This is a theoretical argument, but I’d argue it’s a well-grounded one.”

Here’s the relevant bit, all emphases in the original:

I obviously can’t prove this would make a difference in a case like the IAT’s premature rollout and celebration. This isn’t the sort of thing you can submit to a randomized controlled trial. What I can say, as a journalist who has covered many controversies at the intersection of science and public policy for almost a decade and a half, is that the prevailing ideologies of scientists (and the journalists who cover their work) certainly seem to matter a great deal. Why wouldn’t they? What possible case is there to make that ideological conformity doesn’t unduly lubricate the path from the conception of an idea to its publication and real-world application? Does anyone actually think that the IAT would have been disseminated as quickly if it had been born in a milieu of genuine viewpoint diversity?

I need to be careful here, because I am speculating in the absence of data. It’s also important to note that ideas get adopted prematurely not just because of ideological conformity, but also for other, more basic reasons: If you’re a young scholar in any field, you’re going to feel a gravitational pull toward whatever the currently dominant paradigms are. When implicit bias was the dominant paradigm in discrimination research, young social psychologists were incentivized to study it and — if they were lucky — to produce a paper or two that incrementally advanced our understanding of the subject without criticizing it too harshly. That’s how you advance in a field: Skeptics tend to have a bad time, particularly young ones who lack the influence and job security to meaningfully criticize incumbent ideas.

But it seems silly, at this juncture, to deny the downsides of viewpoint homogeneity, especially given how poorly the social sciences have fared in recent years, with their multiple replication crises. I know that some researchers have argued that political bias cannot explain these problems(12), but whether or not that’s the case, what conceivable downside could there be to introducing more political diversity to the process, some way or another? Could it possibly make things any worse?

I’m annoyed at myself for how I wrote that last paragraph. I phrased things loosely and opened myself up to bad-faith critiques: “Some way or another? You mean like concentration camps????” But still: I simply don’t think a competent or good-faith reader could arrive at this point in the chapter and claim anything like what Wilson is claiming — among other things, that the chapter reveals a “breathtaking indifference to censorship.” If Wilson wanted to check whether this was a fair accusation, Google would have been a useful tool. “This Is Sickening,” I wrote of the Rümeysa Öztürk case. I also criticized Bari Weiss for letting Marco Rubio off the hook in an interview and not holding him to account for his role in the Öztürk and Mahmoud Khalil cases, which should have been right in her free-speech bailiwick.

But I’m falling into a trap laid by a hack by even defending myself in this manner: Since the chapter never came within light-years of endorsing right-wing assaults on academic freedom, I shouldn’t have to say “See? See?? I don’t believe in that bad stuff!” even if that happens to be demonstrably true. Regardless of anything else I wrote, there was simply no connection between Wilson’s most inflammatory claims about the chapter and the content of the chapter itself.

In addition to reflecting editorial incompetence at IHE (not a single editor paused and said “Hey, should you tell readers what the chapter is about? And do you think there’s textual support for these rather serious claims?”), this makes everyone stupider. Readers are rendered stupider by being misled about what I wrote. Wilson’s own ideological allies are rendered stupider by thinking their adversaries are stupider than they are (and thinking your adversaries are stupider than they are is one of the worst types of stupidity). If there are genuine problems with my chapter, I (and my readers) are rendered stupider by being robbed of a potentially productive critique.

Least importantly — but still annoying — I’m rendered stupider by having devoted even a small chunk of an otherwise beautiful Saturday to this.

Questions? Comments? Ideas for better uses of my time? I’m at singalminded@gmail.com.

Viewed from a certain distance, all partisan brains have similar tendencies. They might claim to hold not only different beliefs, but diametrically opposed beliefs, but if you fervently believe in something — whatever that thing is — it puts you at risk of certain blind spots.

This is useful to point out because partisans often think that they and their side have unique access to the truth. Through their superior intellect, their uniquely visceral personal experience, or whatever else, they have transcended the limitations of our mere squishy head-computers and have gained access to certain empirical and moral truths that elude the rest of us.

One way of expressing this is via “horseshoe theory,” a term you have probably heard, which Wikipedia describes as the belief that “advocates of the far-left and the far-right, rather than being at opposite and opposing ends of a linear continuum of the political spectrum, closely resemble each other, analogous to the way that the opposite ends of a horseshoe are close together.”

And of course, far greater thinkers than I have been pointing this out in different ways for a long time. “In pre-Hitlerian Germany it was often a toss-up whether a restless youth would join the Communists or the Nazis,” writes Eric Hoffer in The True Believer. Or, more generally and pointedly:

Though they seem at opposite poles, fanatics of all kinds are actually crowded together at one end. It is the fanatic and the moderate who are poles apart and never meet. The fanatics of various hues eye each other with suspicion and are ready to fly at each other’s throat. But they are neighbors and almost of one family. They hate each other with the hatred of brothers. They are as far apart and close together as Saul and Paul. And it is easier for a fanatic Communist to be converted to fascism, chauvinism, or Catholicism than to become a sober liberal.

I had thought William Shirer also had a passage about this in The Rise and Fall of the Third Reich — about the ease with which true-believer Nazi and USSR officials got along — but when I tried to find it all I found was a section on Joachim von Ribbentrop’s fateful trip to Moscow that wasn’t quite what I recalled. When I asked X, I was also pointed to an astonishing passage about that same trip in Paul Johnson’s Modern Times: The World from the Twenties to the Nineties, which I haven’t read:

That night of 23-4 August there was a gruesome junket in the Kremlin. Ribbentrop reported: ‘It felt like being among old party comrades.’ He was as much at ease in the Kremlin, he added, ‘as among my old Nazi friends’. Stalin toasted Hitler, and said he ‘knew how much the German people loved the Führer’. There were brutal jokes about the Anti-Comintern Pact, now dead, which both sides agreed had been meant simply to impress the City of London and ‘English shopkeepers’. There was the sudden discovery of a community of aims, methods, manners and, above all, of morals. As the tipsy killers lurched about the room, fumblingly hugging each other, they resembled nothing so much as a congregation of rival gangsters, who had fought each other before, and might do so again, but were essentially in the same racket.

That’s just… (hopefully) interesting background. But I think a version of horseshoe theory also holds in much lower-stakes situations involving normal everyday people with fairly standard belief systems.

At the moment, a bunch of people on X are mad at a New York Times headline about Ayman Mohamad Ghazali, the Michigan man who attacked a synagogue. At one point the headline read “The Michigan Synagogue Attacker Was a Quiet Restaurant Worker” (presently: “A Community Struggles to Understand Why Their Neighbor Attacked Synagogue”).

Ghazali’s story is complicated. His attack appears to have occurred as a direct response to the death of two of his brothers and two of their children as a result of Israel’s bombardment of Lebanon. But that bombardment, according to the Times, was a reprisal: After the Iran-Israel/U.S. war kicked off, Hezbollah (a proxy of Iran’s) began firing rockets into Israel.

Ghazali sought to murder Jews in a synagogue — a synagogue that, at that moment, was occupied by a large number of children — and no reasonable reader could accuse the article of soft-pedaling what he did. The article unflinchingly relates the events in question:

On the morning of March 12, Mr. Ghazali drove a truck through the door at Temple Israel in West Bloomfield Township, moved down a hallway, and exchanged gunfire with security guards. He was killed. One guard was injured. Roughly 140 children and staff members at the temple’s preschool were safely evacuated.

Before that, yes, there is some color: a scene from a mosque service memorializing a group that included Ghazali’s relatives, and a quote from the imam there recalling having met Ghazali there for what was apparently his first-ever visit to that mosque.

The accusation is familiar: The Times, by choosing this headline, is somehow seeking to excuse or soft-pedal or justify what Ghazali did. A fairer read is that the Times is attempting to explain it.

A few weeks ago, there was at least some similar uproar over the headline of the obituary for Ayatollah Ali Khamenei after he was killed by a joint United States-Israel effort:

To End Wokeness’s target, an outrage-thirsty audience, the headline is so obviously heinous that no explanation of its heinousness is even required: “This is real.” While End Wokeness is a hyperpartisan account whose entire raison d’être is fomenting outrage, here he (I feel like this account has to be run by a man??) is making the same implication as the present critics of the “Quiet Restaurant Worker” headline: endorsement that the Times headline is refusing to accurately describe Khamenei and the evil he represented.

Both of these mini-outrages are quite similar to a series of mini-outrages fomented by progressives over the last decades, including a 2017 pileon over a profile of Midwestern Nazis that grew large enough the Times appended a quick note to the top linking to a pair of published responses.

Here’s how William Hughes of The Onion’s AV Club described that controversy at the time:

The New York Times briefly switched its motto to “All The News That’s Fit To Piss People Off” today, earning widespread ire for a profile on an Ohio white supremacist that seemed more happy to praise his good manners than condemn his hateful rhetoric. Responses to “In America’s Heartland, the Nazi Sympathizer Next Door” — in which reporter Richard Fausset sits down with his subject at both a Panera Bread and an Applebee’s, those two bastions of suburban fast-food normalcy — have been swift and angry, accusing the Times of running a piece about the dangers of normalizing Nazis that actively normalizes a Nazi in the process.

I’m not sure how to interpret this critique — and Hughes was one of many journalists to level it — other than that the critics believe that the average reader has profoundly dangerous cognitive limitations and can be quite easily bamboozled by familiar references: Look, on the one hand, I’m against his beliefs about killing all the Jews, but on the other. . . he goes to Panera and I do like their soup-in-a-bread-bowl combo, so I’m going to hear him out. I know it sounds like I’m caricaturing this view, but what else could Hughes be saying other than this? Isn’t that basically it?

Read more

There were a lot of responses to my Monday piece about (what I consider to be) artificial intelligence denialism on the left. I want to address some of them.

But first: I didn’t mean to imply that librarians just search their catalogs or Google stuff. I wrote too quickly and flippantly. In fact, I made the same mistake I think Emily Bender and Alex Hanna are making: They are letting their moral outrage about AI color their assessment of its utility. The tail is wagging the dog.

Here, because I thought Bender’s argument that Robert Wright should consult a librarian rather than use AI to develop a first-pass understanding of why Rome fell was so silly (okay, technically the word I used was asinine), I then went a bit oversimplistic in my description of what librarians do. A good librarian wouldn’t just type “Rome” into a catalog and be like THERE CHECK OUT THOSE BOOKS — they’d be able to offer a more tailored set of suggestions, in part depending on who they were talking to (professional journalist versus high school student), how in-depth the person was seeking to get, and other considerations. I don’t think any of this derails my main points about why this was a silly suggestion but, just for the record, I apologize to librarians.

On to the responses. A lot of them were thoughtful and I only have time today to reply to a few. Let’s start with a snarky one: “The perfect Singal post,” wrote a Substack user named StPaulite. “Here is a titanic development in the political economy, with trillions at stake, directed by the most powerful people alive. What’s important to know? The left isn’t talking about it the right way. That’s the problem.”

To be fair, I’m not necessarily sure academics have all that much power to steer the development of big events. But if we are going to have people meaningfully researching and theorizing and critiquing (what I believe will be) an extremely disruptive economic meteor, who else do you suggest? It’s lefties who are most concerned with the concentration of power and unfair hoarding of resources, so I think they should take these events seriously.

Here’s an exchange I had with Osita Nwanevu (lightly cleaned up), which was kicked off by his quote-repost of my piece:

Nwanevu (quote-reposting): The question of whether LLMs can actually think is only a pointless distraction if you assume the answer has no bearing upon what and how many jobs they can plausibly replace, or how many individuals and firms will be willing to use them. It materially matters and obviously so, which is why AI firms insist their models are intelligent and steps toward superintelligence in the first place.

Me: So much leans on your definition of “actually think”. . . when you typed that, which specific definition did you have in mind?

And to be clear, I don’t think it’s a pointless distraction in some universal sense — I think AI raises utterly fascinating questions. But yes, I do think in terms of its economic impact — whether it can simulate being a living assistant who completes various tasks competently is what matters most, regardless of what is going on “under the hood.”

Nwanevu: I have no specific definition, which is why I think those boring philosophy of mind debates are important! I don’t think most people do, and yet it seems clear to me that there are material implications to how we think about what it’s actually doing under the hood. Because when we ask whether it can actually think or reason, we are also functionally asking, “Can it be a tool for doctors or a doctor?” “Can it be an artist’s assistant or an artist?” “Can it be a reference for intelligence analysts or an analyst?” Etc. That shifts the character of the economic disruptions and the number and nature of the job losses we’re talking about.

My (limited) understanding of the situation is that modern AI systems are going to remain more or less black boxes for a very long time, and maybe forever. An LLM from this generation, for example, consists of about a trillion “weights” (numerical values that help determine how it processes information, to oversimplify) and responds to your input by multiplying 12,000-ish-dimensional vectors, over and over and over. There’s a sheer level of mathiness here that makes it extremely hard for humans to actually understand, in a detailed way, what’s going on under the hood. That’s part of what the Gideon Lewis-Kraus New Yorker article I touted was about, and part of why researchers are often surprised by what their own creations are capable (or incapable) of.

Generally speaking, I think there have been surprises about AI’s capabilities in both directions — it’s better at certain things than experts thought, and in some areas has lagged behind where those experts hoped we’d be by 2026. The general trajectory, though, has been rapid improvement, especially lately. For my purposes, what matters is the number and nature of tasks that AIs have achieved what some wonks call human parity: that is, they can compete with paid humans at completing that task. This is not necessarily the same as being able to do a job (which consists of both many tasks and other forms of executive functioning), but it’s obviously an important benchmark and the first step toward full-blown job replacement. It seems undeniable that in just the last two years, AI has reached human parity and in some cases significantly exceeded it at various types of coding, copy-editing, translation, summarization, image generation, and other situations.

Here’s a photorealistic image of “an ape protecting its favorite toaster. the toaster has been through some stuff. they are in the high desert of texas.”

That took less than 20 seconds for ChatGPT to create for me.

Because we get so used to new technology so quickly, I think there are a lot of people who don’t understand how remarkable this is, how much of an improvement it is over the technology we had two years ago, the extent to which it represents improvement in areas of computing that were frustratingly stagnant for a very long time — and how threatening this is to human labor.

It’s threatening to human labor because there are so many other tasks where we’ve seen recent, explosive improvement. Surely you’ve had an infuriating encounter or 10 with those voice-automated customer-service systems? They’ve been the subject of hackish stand-up routines for what, 10 years? Twenty? Have you talked to one of those systems recently? They aren’t perfect, and in fact during my last interaction with one, belonging to a car rental company, it gave me false information. But there has been clear, obvious, holy shit improvement. They are about as fluent as a human tech-support employee.

That’s why I’m skeptical of Nwanevu’s framing, and that’s why I raised the point, in my last post, about drinking in a bar and discussing the nature of intelligence versus actually using these systems. If I broke down the ape image-generation task to its constituent parts, it would sound like something that couldn’t possibly generate a photorealistic image of an ape guarding its favorite toaster in the Texas high desert. “Whatever that is, it isn’t thinking!” you’d say. Maybe you’ll add “It definitely isn’t creating,” a bit haughtily. Nwanevu’s question, I think, is geared at generating some definition of actually think that will “prove” AIs will be incapable of this, that, or the other thing.

But my argument, over and over and over, is that if we’re talking about the impact all of this will have on human labor and human society, the only meaningful answer here is: It doesn’t matter whether “the computer” is actually thinking, actually creating, or can in any sense be said to “understand” what an ape, a toaster, or the Texas high desert is. It doesn’t matter. What matters is the usefulness of this technology, and that it has reached or exceeded human parity at an ever-growing number of tasks that used to help people make a living. To repeat myself: At a certain point, a new invention is so obviously, indisputably useful — that’s the only claim I’m making, useful — that it requires sticking your head deep into the sand to deny that it is going to have a major impact on society.

Which brings us to Freddie deBoer’s response, which, true to FdB form, was also accompanied by a separate, somewhat vituperative email. But here’s what he posted publicly, slightly cleaned up. There’s a lot here, so I’ll bold the few points I’m going to respond to directly.

What’s denialism? Is pointing out that high-profile examples like protein folding have hit a massive wall denialism? Is it denialism to point out that AI company ghouls like Dario Amodei keep making outlandish short-term predictions that keep not coming to fruition? Is it denialism to point out that it’s increasingly clear that LLMs can’t be built without inducing repetitive and serious hallucinations, hallucinations that make their deployment in mission-critical situations potentially impossible? Is it denialism to point out that even if we take the most outsized claims about what AI is theoretically capable of seriously, the entire history of technology demonstrates that seemingly minor implementation issues can totally wreck projections about the adoption and use of a given technology? Is it denialism to point out the the widely predicted and much ballyhooed claims of massive productivity and GDP growth stemming from the internet never ever arrived, and that the 2020s have seen another half decade of anemic growth compared to the mid-20th century, and that none of the promise of productivity or GDP gains show up in the data whatsoever? Is it denialism to point out that even the most enthusiastic proponents now admit that the scaling “laws” are broken? Is it denialism to point out that LLMs cannot continue to be fed more data at similar scales because there simply has not been anything close to as much written language produced in the history of the world that would be required for that to happen? Is it denialism to point out that people keep declaring problems like machine translation over, only for end-user solutions like Apple translation via AirPods to keep failing in real-world applications? Is it denialism to point out that every single time new technologies have been predicted to result in massive job losses, over time, any short-term losses have been overcome and in fact have led to greater job gains, repetitively, over and over again throughout the history of modern technology? Is it denialism to point out that saying “this time it’s different” is exactly what people said about the loom, the printing press, the telegraph, the computer, or the internet, and yet in every case those technologies eventually caused job growth rather than job loss? Is it denialism to point out that the history of futurism is a history of immense failure and wrong predictions, again repetitively, over the course of centuries, and that the research tells us that human beings have not gotten better at futurism and prediction over time? Is it worth saying that, again, every single person of every generation that has predicted sudden immense technological change has felt sure that this time it was different? Is it denialism to point out that the future almost always ends up being very, very similar to the recent past, and that there’s no reason to think that we’re among this weird blessed generation who gets to experience this sudden tremendous change? Is it denialism to say that extraordinary claims require extraordinary evidence? Is it denialism to point out that your consciousness system works overtime to convince you that you live in a very special time, even though definitionally no pressure of time is transcendently special and that all of human life eventually collapses into the mundane? Really? That’s all pure denialism?

I have no idea what’s “lefty” about any of that, and frankly the word “denialism” now just seems to mean “opinion held by anyone that is not identical to my own when it comes to AI.”

For each of these I’m going to respond not only to the claim itself but to what I view as the style of silly AI-denialism reasoning underpinning it.

“Is it denialism to point out that it’s increasingly clear that LLMs can’t be built without inducing repetitive and serious hallucinations, hallucinations that make their deployment in mission-critical situations potentially impossible?”

The basic move here is to endlessly shift the goalposts. LLMs hallucinate, yes. Perhaps they always will (as some experts have argued). But it’s very easy to train people to be on guard for these hallucinations, in the same way you can train people never to cite Wikipedia directly. As for “mission-critical situations,” that’s obviously very important in a maybe-it-will-kill-us-all sense, but it also doesn’t defeat my argument at all. If AI progress froze right where it is now, it would end up having a hugely disruptive effect once hirers fully understood how, why, and when to use it. We don’t need to bring terrifying autonomous ICBM systems into this.

“[S]eemingly minor implementation issues can totally wreck projections about the adoption and use of a given technology.”

A lot of things could happen! Surely some areas of AI will flame out, in some cases before they even get to public users. Again, how does this defeat what I’m saying?

“[P]eople keep declaring problems like machine translation over, only for end-user solutions like Apple translation via AirPods to keep failing in real-world applications.”

Everyone has to watch or rewatch Louis CK’s famous airplane bit on Conan before reading another word of my nonsense, and if you haven’t seen it, I can’t be held responsible for the references you won’t get:

Give it a second! It’s going to space! Can you give it a second to get back from space?

This is my favorite one because it really highlights the amazement-treadmill aspect of this. Until recently, the idea of real-time translation via earbuds was strictly sci-fi. It was something that, while by no means impossible in theory, would be very hard to pull off in practice, especially in an affordable way. Now we have it — because of advances in AI — and Freddie deBoer believes it has “fail[ed].” Has it? If you want, you can check out “Airpods Pro 3 Live Translation Demo & Review” from five months ago. The reviewer indeed is wearing tiny little earbuds in his ears linked to his phone that can, in fact, translate spoken language in real time.

The problem is. . . there’s a bit of a delay.

Oh, really? What happened next? Did you fly through the air, incredibly, like a bird? Did you partake in the miracle of human flight, you noncontributing zero?

I want to make sure I’m not misunderstood here: I am not some slop hog for the latest Apple or Google or Meta invention. I have never owned a VR headset, I still have an Android phone despite how much it sucks and destroys my group texts, and I haven’t and never will be an “early adopter.” My point is not that we should trust tech companies when they tell us they’ve created something amazing. My point is that when they do create something amazing — and something that seems very likely to have real-world economic consequences — we should acknowledge that rather than shifting the goalposts. “Oh yeah, it translates Spanish in near real-time? Well. . . I don’t see it manifesting a hot Puerto Rican girlfriend for me out of thin air, so what use is it?”

This is silly.

Finally:

“Is it denialism to point out that every single time new technologies have been predicted to result[ ] in massive job losses, over time, any short-term losses have been overcome and in fact have led to greater job gains, repetitively, over and over again throughout the history of modern technology?”

This is probably deBoer’s strongest point, and where I am on the weakest ground due to my lack of any firm grounding in economics. So I’ll proceed cautiously.

A lot is leaning on “over time” here. Let’s stick to the United States for simplicity’s sake. The Great Recession and Covid were both economically devastating. They both left giant scars on people’s material well-being and had strange, largely deleterious impacts on our politics. Speaking quantitatively, we “bounced back” from both, but that doesn’t really undo the damage to individual victims of these events, or to our “national fabric” or whatever you want to call it.

If I’m right and AI is going to cause some sizable shock, we should be concerned about that. “Eventually it’ll work out” isn’t really an answer, especially because things seem rather precarious at the moment.

I am generally a believer in the idea that humans have made real progress over time. I think it’s silly to ignore the fact that fewer babies die, people live longer, and there’s less suffering in general. That being said, 2026 America is a fragile place. We have a historically bizarre and authoritarian president who represents a party that is, even in the best of times, opposed to meaningful redistribution; an absolutely pathetic, feckless, weakling, loser opposition party; and we’re still living in the shadow not only of Covid and the Great Recession but of forms of economic dislocation and hollowing-out that preceded all that by decades. If any developed country and its political system is ill-equipped for a new economic shock, it’s ours. I think there are a lot of reasons to be worried; I think we’re a once-proud prizefighter swaying like a drunk in the seventh round.

And that’s just if AI causes a regular economic shock. I’m not going to pretend I’m deep enough in the technical weeds to know whether some of the more ambitious potential capabilities of AI are actually going to come to pass. Will it actually be able to improve itself and consistently invent new, useful things, rather than “just” achieve human parity at preexisting, well-understood tasks? Given how much is riding on this for the giant firms with which our economy is hopelessly entangled, it wouldn’t surprise me. And if these slightly more far-out realities come to pass, it could simply be that this represents a new sort of technological advancement we have no real blueprint for, that even the best economist can’t meaningfully forecast.

Some Reader Responses Worth Highlighting

Read more

Modern artificial intelligence technology seems to already be having a major impact on at least one sector of the labor market: software development. Because AI systems are now — at least by certain measures — better and quicker coder than humans, company after company after company have announced significant layoffs. People who have trained their entire lives to develop software for a living are publicly expressing panic and existential dread. More anecdotally, I recently met someone who works at one of the giants out here in the Bay and she told me her employer explicitly told her that due to the new tools available, they expected her to increase her productivity by a specific, set multiple. (I asked her if the company could actually track her productivity with that much precision and she shrugged and said, Sorta.)

I believe AI is going to have a major disruptive impact for these reasons, but also because of my own experiences. I use AI every day as an utterly fascinated but also skeptical user. I’m utterly fascinated because, while I am not a Tech Guy, I’ve been using computers for three and a half decades — I was fascinated with computers as soon as I was first exposed to them, which led to my parents buying me a 386 — and am intimately familiar with the sorts of tasks they’re good at versus bad at. There has been an explosion in AI’s ability to do stuff it used to be bad at. Recently I had Claude make a PowerPoint and a simple JavaScript application, both for live shows, and it was simply insane: Of course there were minor rough spots, but the experience was indistinguishable from hiring someone for those tasks, going back and forth via email over edits and revisions, and so on. These are not super fancy design or software development tasks, but they are tasks I would have had to previously pay for or mooch off a friend for. If it is replacing human labor for those sorts of tasks, how could it not be replacing somewhat more advanced tasks? Of course it is. Hence the layoffs.

But I’ve been far more influenced and blown away by my normal, daily use of AI. At this point I use AI — mostly Claude — as something of an extra research assistant and set of “eyes.” It can quickly pull together a summary of some controversy, and as long as I follow long-ingrained Wikipedia Rules (take the text itself with a grain of salt and focus more on the sources it provides), it’s exceptionally useful. I also use Claude as an extra set of “eyes,” in the sense that if I input a draft of something I’m working on I can get perfectly competent — and steadily improving — editorial feedback. That doesn’t mean it’s perfect. In fact, if someone with almost no experience writing or reporting drafted a 1,200-word column and then trusted Claude to do the brunt of the legwork cleaning it up, I think the results could be potentially disastrous. In much the same way I read a human editor’s suggestions in an open-minded but skeptical way, that’s how I treat Claude’s. Sometimes I agree with the suggested edits and sometimes I don’t, and it is largely useful to me because I have a lot of experience writing and editing and can distinguish good from foolhardy feedback.

Overall, though, I am shocked at how good it is. Whereas with software development I am taking experts at their word, here I’m not: I have been doing this (writing and editing) for a long time. I have had good editors and bad editors. Claude is a pretty good editor that is effectively free ($20/month for the version I use) and that can read a 1,200-word draft and provide generally good, generally incisive feedback in about four seconds. Anyone who doesn’t think this is revolutionary technology is either being intentionally obtuse or lacks the sort of expertise required to render this judgment in the first place.

It strikes me as undeniably true that AI is going to eat a lot of jobs. I don’t think you need to be a labor economist to realize this — I think you just have to have a lot of experience using it and knowing about the sorts of tasks it is getting better at.

An analogy: Imagine, somehow, you’ve never seen a car smaller than a bus and you travel regularly between San Francisco and Los Angeles. You are at the whim of airlines and buses and the patchy rail service in this corridor. One day you find out about cars. You also find out that while cars are pretty expensive, they’re cheap enough that most adult Americans can afford to at least buy a used one.

You: Wait a minute. . . a machine that whisks you wherever you want, when you want, along any route with roads, at a relatively high speed? Why didn’t anyone tell me about this? Isn’t this going to revolutionize everything?

Smug, skeptical friend: Ummm, not so fast. Are you really qualified to say that when you aren’t a transportation or urban planning expert?

At a certain point, a new invention is so obviously, indisputably useful — that’s the only claim I’m making, useful — that it requires sticking your head deep into the sand to deny that it is going to have a major impact on society.

Unfortunately, there’s a fair bit about that going on. And it’s centered on the left, which is where we’re going to need the strongest, smartest critiques of AI. After all, these technologies are owned by unfathomably huge, juggernaut corporations. Those unfathomably huge, juggernaut corporations are already getting in bed with unfathomably huge, powerful governments (you may have seen that there’s been a bit of news on this front lately), and all of this is going to raise fundamental questions about fairness, what governments owe their people, and social welfare in general. That is, questions that the left is supposed to be concerned with.

It’s a really bad time for lefty thinkers to be asleep on the job, in other words.

Speaking of Which

I’ll get to a specific, high-profile example in a minute, but broadly speaking, lefty social-media discourse on AI involves a huge amount of abstract philosophizing. You’ll see sweeping claims that AI can’t do this or that — usually that it can’t “think” or “reason.” These are profoundly interesting debates if you’re a philosopher and profoundly boring debates if you’re interested in AI’s real-world impact on society.

I don’t even know how to find anything useful in the claim that Claude can’t think or can’t reason when I send it a draft of something I’m working on. I understand what’s being claimed — that the process by which Claude generates its output differs, in key ways, from the process by which a human editor would read a draft and generate his or her suggestions. But this doesn’t really matter. Either the end user finds the service useful in this context, or they don’t, and it is wildly useful. If you’re offended or think I’m anthropomorphizing, I’ll use scare quotes to soften the blow. Claude can clearly “read” and “understand” and “critique” and “evaluate” and “edit” and “offer suggestions to improve” anything I draft. For each of these terms, just swap “simulate the responses of a competent human reader.”

If you were a philosopher of mind and we were at a bar with several empty pint glasses before us, I’d have a ton of questions for you about all of this. It would be a super interesting conversation about the very nature of thinking and writing and communicating. But you’re not a philosopher of mind (well, most of you aren’t) and we’re not at a bar. It’s Monday at noon and I’m at my desk, writing, and I have a ton of work to do and, as always, I could use some extra help. Claude is extremely helpful. I still rely on and in many ways prefer human help — I continue to employ a wonderful human copy editor and am about to hire a human researcher — but Claude is extremely helpful.

There will be losers in all of this. A lot of them. I think my own jobs are fairly down the list of threatened ones, but I could be quickly and disastrously proven wrong about that. So what smart lefty thinkers should be doing now — to the extent they can influence the situation at all — is figuring out how to prevent this technological revolution from immiserating the people left behind.

Instead, there’s a disproportionate amount of moralizing and philosophical “bullshit” — I call it bullshit not because the underlying debates are uninteresting, but because they don’t matter if the end user finds the product useful.

There was a long conversation on this subject I found particularly telling (and vexing). It was on the journalist Robert Wright’s podcast/show NonZero. He interviewed Alex Hanna and Emily Bender, co-authors of The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want, which came out last year. Hanna and Bender are both celebrated scholars on this subject. Bender co-authored the famous 2021 “stochastic parrots” paper, which is overdue for an in-depth revisit.

I have not read their book. I trust Bentham’s Bulldog, I read his review, and unless I plan on writing about The AI Con further, I’m probably not going to read it, to be honest. I think it’s perfectly fair to judge a pair of celebrated academics on the basis of how they present themselves during an hourlong interview podcast. And they — particularly Bender — came across as condescending, pedantic, and bizarrely out of touch with the on-the-ground reality of AI, given that this is supposedly their area of expertise.

Let me try to lay out their argument as faithfully as possible. Early on, Bender criticizes “this idea that AI is a thing, it’s inevitable, it’s imminent,” and contrasts it with her and Hanna’s “concerns not about anything AI would do but about what’s being done in the name of artificial intelligence as people try to chase this fantasy, the harms that are happening to actual people and the planet.”

Here and there, Bender and Hanna allow that AI is good at this or that task, but mostly they focus on the idea that the entire concept is itself incoherent, a con, or both. There is, of course, AI hype, and it can be hard for nonexperts such as myself to mistake hype from reality. But every transformative piece of technology also brings with it hypesters and fraudsters, and the existence of hype, even a lot of it, doesn’t mean much on its own.

Hanna and Bender don’t really think AI can “do” anything (I’ll return to that soon). They also don’t even seem to like using the term AI as one unified concept. They have lots of preferences about subtle matters of word choice in general. One of the stranger and more annoying exchanges occurs when, after Wright asks, “So where do you think it’s helpful?” Bender responds, “What do you mean by it?” Wright, who elsewhere in the interview exhibits clear familiarity with the underlying technology, does respond a bit tautologically, but this is sort of a silly question in context: By it he just means tools like ChatGPT, Claude, and so on!

Hanna then allows that something like ChatGPT might be useful for an independent journalist like Wright that “an LLM may be useful in so far as structuring your blog posts and then doing the verification after that to ensure that everything is in accordance with what you would want to say.” Wright tries to interject to clarify that he doesn’t use it to structure his posts (I don’t either, and I actually don’t think the vast majority of professional writers would), mentions that he might sometimes use it to “verify” something, and then Bender breaks in, “Wait, you’re using ChatGPT to verify something?”

Wright: Well, I’m more inclined to use Perplexity because it includes links, but I would say all of them are more and more inclined to include actual links to sources. But actually, not just verification. The most valuable thing is actually the research phase, and ultimately, you can use it to find the actual sources. And so, yeah—

Bender: So I’m saying, do you always go back to the actual sources? Or do you ever take just what comes out of Perplexity?

Wright: I go back to the sources if I’m going to report a fact or a quote or something like that, yeah.

Bender: Yeah, okay, good, because I was gonna say otherwise, you’re not doing journalism if you are not going back to the original sources, because these things make paper-mache out of their input. And yes, sometimes a link comes out, and it might be an effective way to get to that link for you, but — so you’re talking about the use case of basically a search engine. It’s a search engine that allows you to use more natural language rather than keywords. It’s a search engine that maybe does an end-run around some of the enshittification of search, right?

What jumped out at me, in relistening to this segment, is that Bender and Hanna don’t seem particularly interested in actually letting Wright explain exactly how he uses AI. On the one hand, fair enough — they are the guests and he is the host. But they just sort of put a lot of words in his mouth. Bender latches on to this idea that he is basically using it “as a search engine,” which seems like a real oversimplification of what Wright is saying. Then she explains that using it as a search engine is not a good idea.

Maybe I’m projecting, but I think Wright starts to sound a little worn down by this point:

Wright: I’m curious: Why is it not a good idea?

Bender: So most people who use a chatbot as a search engine are not going to actually use [it] as a search engine. They’re going to use it as an answer engine. They’re not going to drill down and look at the sources, and so they’re not going to see how those answers are located in the information ecosystem and how the sources speak to each other. As a trained journalist, I’m willing to believe that you are above that, and that you always go to the source, but the average user doesn’t. And then on top of that, you’re talking about systems that are environmentally ruinous. You’re talking about systems that are built on labor exploitation, and you’re talking about systems that are built on data theft. So to the extent that you are getting value out of that, compared to the old-fashioned search engine, compared to connecting with people and asking questions, that value is coming, not for free and not for whatever your monthly fee is, but actually at a very large cost to society.

I am not an expert on the power question, but just so that you are all aware that there is another side to this argument, here’s what Bentham’s Bulldog says in his review:

The authors then complain about the climate impacts of AI. They do this mostly by just citing the raw amount of energy used, and comparing it to other things. But it’s not surprising that a technology used by billions of people uses lots of power. If you prompt an AI 100 times, that will use about .1% of your daily power use. Compared to other activities we perform regularly, like driving, AI uses relatively little power (see here for more). Predictably, they also repeat the even more bogus water use canard — eating a single hamburger is over a million times worse for water usage than prompting AI.

I have not checked these statistics. I will say that I am biased, because I am supremely skeptical of the idea that using an LLM brings with it “a very large cost to society” that I should care about, relatively speaking. That is, as someone who sometimes drives a car when I don’t, strictly speaking, “have to,” and who consumes dairy on the regular, and who probably — okay, definitely — has food delivered to me more than I should — the idea that it’s my chatbot usage I should feel bad about. . . it just doesn’t pass the smell test.

That’s especially true when you at least recognize the possibility of chatbots saving energy over nearby alternatives. Compare two situations: A 2022 (that is, pre–powerful-chatbots) Google search for “reasons for end of Roman Empire,” which will entail a ton of follow-up clicks and searches, versus what I just did, which is prompt Claude: “Could you write up two thoroughly cited paragraphs reflecting contemporary thinking on the reasons the Roman Empire fell? Do this in a way that will be useful for someone like me who is a total novice to the subject and wants to dive into the existing literature.” I’ll stick the output in a footnote1, but I’d argue we should at least consider the possibility that this is a more energy-efficient way of getting the user that basic rundown and initial list of links than the technology we had four years ago.

I didn’t pick that example at random:

Wright: I mean, what I’ve found valuable is, if there’s something I just know nothing about — I mean, like, theories about why Rome fell or something — I find it a much more efficient way to just get a sense for the landscape, because, you know, it’s like having an actual expert that you can interrogate, and—

Bender [like Hanna, shaking her head vigorously]: No, no, there’s no expert there, and you’re not getting a sense of the landscape.

Wright: I don’t know — it’s closer than any other option I have. I definitely can’t summon—

Bender: Oh, go to the library! Talk to a librarian!

Wright: Yeah, but it’s so much slower! It’s so much slower!

Bender: Oh dear, you have to go out in the world and talk to people, like, you’re missing out on so much there, and on top of that you’re getting a fake view of a fake landscape.

This is such a striking and revealing exchange.

First, while I have nothing against libraries or librarians, and in fact quite like working in libraries, it is ridiculous, in 2026, to tell someone in this hypothetical situation to start with a librarian. If I ask a librarian for information about why the Roman Empire fell, s/he will either query the library’s catalog for certain keywords, or Google the question. The prior option will mean I’m at the mercy of that particular library’s collection of Rome-related materials, while the latter will just bring us back to algorithms and, increasingly, AI (given that that’s what pops up at the top if you Google something these days). And yes, as Wright indicates, that is a much slower process.

I’m not sure what Bender means by “fake view of a fake landscape.” If she’s talking about the possibility of hallucinations, that’s a problem, for sure, but 1) less and less so with each new model, and 2) a problem you can avoid by maintaining Wikipedia reader–level savviness and double-checking any specific claim you see in the text. If she’s talking about bias in a broader sense — that is, the literature itself is biased in favor of some explanation for Rome’s fall — then how would a librarian magically overcome that bias? I’m confused and wish Wright had pressed her on this (in his defense, there was a lot of cross-talk at that moment and at this point they had hit the hour mark).

Knowing Without Using

I think Bender and Hanna, and particularly Bender, really hate this technology. They aren’t shy about it and that’s probably the neatest explanation for how you can get to the point where, with a straight face, you’re advising a professional journalist to seek a librarian’s help getting started on a research project in 2026. One key fact is buried deep in the interview and actually comes right before the Rome exchange, after Bender allows that in certain specific cases, she does think LLMs engage in “understanding”:

Bender: But the problem is, it’s all dressed up in all of this other fluff, which plays on my ability to make sense of language and makes it seem like it’s not just a machine that has been able to map what I said to some specific routine to carry out, but rather some thinking entity that has any right to use I/me pronouns that I’m interacting with.

Wright: Mm-hmmm. You know, I was going to ask you this question. You said you never go to ChatGPT. Do you not use the chatbots, the LLMs yourselves in your lives—

Bender: None of them.

Wright: —and if you do use them, how do you use them? But do you not use them at all?

Bender: I do my very best to avoid reading any synthetic text or consuming any synthetic media. I am not interested in reading something that nobody said, that nobody wrote.

Hanna: I just check how bad they are at things. [Hanna then explains that she keeps hearing it’s good at coding but remains unconvinced.]

I’m confused how Bender can make such confident claims about LLMs — especially her lack of concern that they will actually take jobs — when she makes a point of not using them. This might explain why my trajectory has been the opposite of hers. The more I use LLMs, the more impressed I am, the more useful I find them, and the more worried I am about their economic and more general societal impact. Why can’t we hold all those ideas in our head at the same time?

At the opposite end of Hanna and Bender is the New Yorker’s Gideon Lewis-Kraus, who wrote a fantastic article about Anthropic’s attempts to understand this strange entity it has built, Claude. He then went on PJ Vogt’s podcast Search Engine to talk about his piece.

One part at the end continues to jump out at me, especially after having listened to the NonZero interview multiple times. “I think people wanna have one feeling about this stuff,” he explains at around 44:45. “They wanna be angry about it, or they wanna be messianic about it.” But in the course of reporting on this story, Lewis-Kraus gave up on the idea that there’s any one way people should feel about AI — he just thinks the technology is too big and consequential for that to be the correct approach. “We have this unprecedented entity that is the only other thing besides us that can talk,” he explains. And “There’s nothing that it doesn’t touch on.”

In context, by talk Lewis-Kraus means engage in an intelligent, sustained conversation. And he’s right! Humans have created a type of software that you can talk to for hours. It’s incredible. It’s going to have all sorts of effects — it’s not easily boiled down to any one outcome. It definitely can’t be written off the way certain people are doing.

In the Search Engine episode, Vogt then explains that contrary to certain negative stereotypes of techbros, according to Lewis-Kraus’s reporting, the folks actually paid to work on these issues at places like Anthropic care deeply about the impact this technology will have on society. But “our intellectual culture is really bad right now”and they are “looking for grown-ups and not really finding people to have a conversation with.”

Lewis-Kraus responds:

And we are playing a role in that too. Every time someone on our side, so to speak, is just like this is all a parlor trick, this is all hype, this is all smoke and mirrors, we are abdicating our own responsibility to be involved in this. . . . Of course [hypesters] exist and of course all of those people are part of this system too, but there are others who want partners and talking about this stuff, and that means that we all need to try to rise to the occasion.

Maybe I’ll close with one other analogy: capitalism. Critics of capitalism can be roughly divided into those who treat it as a fundamentally evil force that destroys everything, full stop, not much more to say, like seriously fuck capitalism, and those who, even if they are harshly critical of it, acknowledge the obvious reality: Capitalism also has upsides and convenience for a lot of people.

Who do you trust to produce better thinking and more tenable solutions to the problems that capitalism, whatever you think about it, does inevitably produce? Which type of critic is better equipped to offer the right sorts of laws and regulations and norms?

There’s something similar going on here. You can be the sort of person who parrots lines about. . . well, parrots, who insists that an in-fact revolutionary and deeply fascinating technology is a parlor trick, who isn’t particularly interested in how normal everyday people use this technology. You can avoid using it entirely yourself because you don’t like the stink of it. You can scoff and condescend and, in the long run, ghettoize yourself out of the conversation.

Or you can recognize what’s really going on and how desperately in need of adult conversations and adult solutions we are. If not now, then very, very soon.

Questions? Comments? Responses to the rumor that I once dated an LLM for six months without realizing it? I’m at singalminded@gmail.com. Image: LONDON, UNITED KINGDOM - 2026/02/28: Protesters gather with anti-AI placards outside OpenAI offices in King's Cross during a march against unregulated Artificial Intelligence (AI) and data centres. (Photo by Vuk Valcic/SOPA Images/LightRocket via Getty Images)

Not the subject of today’s post, but earlier this week I had a column in The New York Times about youth gender medicine and “trust the science” that might be of interest to some of you.

Elon Musk got a lot of attention for a recent exchange he had with the philosopher Amanda Askell, who works at Anthropic and whose job is to help determine the “values” of that company’s AI’s programs (to anthropomorphize a bit).

Responding to a WSJ profile of Askell and her efforts, Musk — singing a familiar tune — said “Those without children lack a stake in the future,” sparking the rest of the exchange:

Askell: I think it depends on how much you care about people in general vs. your own kin. I do intend to have kids, but I still feel like I have a strong personal stake in the future because I care a lot about people thriving, even if they’re not related to me.

Musk: You cannot understand my point until you have a child, anymore than someone who has never experienced true love can understand love.

Musk is engaging in a very stereotypically “woke” pastime known as standpoint epistemology. Standpoint epistemology came out of feminist philosophy. The Stanford Encyclopedia of Philosophy entry for “Feminist Epistemology and Philosophy of Science” is authored by Elizabeth Anderson, one of the most highly regarded public-intellectual philosophers alive today. According to Anderson, this field “identifies how dominant conceptions and practices of knowledge attribution, acquisition, and justification disadvantage women and other subordinated groups, and strives to reform them to serve the interests of these groups.”

At some point I want to better understand this area of philosophy. My sense is there’s a lot of rigorous, legitimate material here worth exploring (philosophers, email me to tell me what you think). I don’t know enough about this subfield to criticize it, so that’s not what this article is about.

Rather, it’s undeniable that legitimate ideas from academia sometimes trickle down to the public internet and get bastardized and adulterated and ruined. Standpoint epistemology (or SE, as I’ll call it) is a particularly annoying example, because, as practiced by people arguing online, it is irredeemably dumb.

*** Traditionally, it was “woke” people who wielded this in the most annoying and anti-intellectual manner, particularly (but not exclusively) online. From, I dunno, call it 2015 to 2021, a stultifying culture of identitarian deference infested many (most?) progressive spaces. That is, it was considered normal and legitimate to defer to the more “marginalized” person in a given conversation, no real argument needed

“Marginalized” usually meant “has an identity label that is historically associated with oppression,” and as result — and I know this is hard to imagine but I promise you it was routine — you frequently got situations where, like, a financially struggling twentysomething writer type was angrily told that they were required to defer to a tenured professor, because that professor was “marginalized” (often meaning something like “a biracial person with a BA from Wesleyan and a PhD from Stanford”). Of course I’m cherry-picking a little here, but my point is the idea of being “marginalized” became completely untethered from anything reasonable, and even if it hadn’t, of course this sort of deference is silly and condescending. You don’t say, “Okay, okay, if you say so! Sorry for disagreeing!” to someone you actually respect.

Identitarian deference created a great deal of annoyance online in the form of countless cartoonishly idiotic exchanges where some person would politely disagree with (for example) a particular black woman about a particular question, and then would get dogpiled by people lecturing them about how morally grotesque it was for them to “not listen to black women,” “talk over black women,” and on and on and on. It was just this constant abstracting of specific conversations, a process where everyone got into a hot-air balloon and took it up to 40,000 feet and looked down and could only see the colors of the characters involved — none of the actual details — and would then evaluate the situation on that basis alone. (Leftists who criticized identitarian deference from early on, like Freddie deBoer in a now-private post (or see his more recent explanation here) and Matt Bruenig as far back as 2013, deserve credit for doing so at a time when this was genuinely risky. Adolph Reed and Walter Benn Michaels are probably the leading old-heads in this camp, and Olúfẹ́mi Táíwò wrote the version of the critique that probably made the most headway among the sorts of leftists who dislike deBoer and Bruenig for being “class reductionist” or whatever.

To be clear, this isn’t how the vast majority of normal people of any color think about these issues — most people are normies on race and identity — but online, where there’s a constant battle for attention and influence and prestige, this kind of dysfunction was completely rampant in progressive spaces. Worse, it did real-world damage to real-world progressive institutions. Ryan Grim’s now-classic 2022 Intercept piece about meltdowns in the progressive world involved cases that often came down to this sort of identitarian deference. Within these institutions, you couldn’t really question accusations of bigotry or sexism due to these norms, which naturally empowered the sorts of people who make lots of accusations, which often made it hard to evaluate their accuracy.

***

The basic SE insight that some members of some marginalized groups might, on average, have access (or easier access) to certain types of knowledge than members of more powerful groups seems pretty obviously defensible. But it takes a wild leap to get from there to the setups that solidified in progressive circles. SE, as practiced by progressives in our online and offline communities, first pretends that its wielders can neatly sort the world into “more” versus “less” marginalized people during an age when this is increasingly hard to do, and then commits a strange version of the “noble savage” fallacy, which holds that the “more” marginalized person somehow has unerring insights into even complex policy questions.

Read more