American ‘woke’-fighter takes Maarten Boudry’s place.

Young people out of work or at risk of entering unemployment after school are set to benefit from a first of its kind AI bootcamp kicking off in the north-west.

One of the peer reviewers who rejected this paper put it best, so we will give them the last word on this thread: >The authors' comments about the MBFC scores reflecting the "common idea of the liberal media" miss the Okkam's [sic] Razor solution that this is in fact the distribution of bias of media outlets.

> One of the reviewers who seemed too close to the issue called it “unseemly” and “inflammatory” to talk about the central role and beliefs of private philanthropists funding open science. https://scatter.wordpress.com/2026/07/31/science-funding-and-the-cherry-picking-problem/

Science publishing–like all of science–requires money. There’s no getting around it, even with the open science movement. What appears free to many authors and readers still requires money and infrastructure. Popularizing the open science movement cost money, too. I tried to talk about where that money came from in my new paper with Molly King, but that part of the paper was censored. One of the reviewers who seemed too close to the issue called it “unseemly” and “inflammatory” to talk about the central role and beliefs of private philanthropists funding open science. After multiple rounds of review where we demonstrated the reviewer was wrong about basic facts such as the names of donors and the amounts of money, they retreated to their core claim: discussion of funders and motives “doesn’t belong” in a paper about feminist ethical considerations and open access publishing.1 Our paper’s acceptance was made conditional on cutting it. I think it’s too important not to share. Two years ago, when we were writing the paper, I worried that the way open science was funded left the door open to far-right, anti-science co-optation. We speculated that it opened the door for wealthy, anti-science, anti-woke interests to undermine science. During the revision process, my fears began coming true. Today, we have a new “gold standard science” initiative coopting science reform rhetoric, and some of the same players we discussed are involved in the Trump administration’s efforts to undermine science. I should start with **some caveats**. This section of the paper was my pet project; my coauthor was along for the ride. This post, too, is my own, although of course we discussed it. I bear no ill will toward the editors or the reviewer who demanded we cut the section. They were very knowledgeable about open science, and the paper is genuinely better as a result of their feedback on other parts of it. Understandably, people get sensitive when someone suggests their work may be funded or cited for nefarious purposes, even if they themselves are virtuous. It is not a pleasant possibility to consider. Indeed, it was “ruining an otherwise reasonable and interesting paper” for the reviewer. Nor is there any purity politics here: I use the same open science infrastructure we wrote about, and I have taken money from billionaires whose politics I don’t agree with. As Philip Cohen rightly pointed out in a bluesky conversation, with the cuts and censorship in federal funding, private philanthropy may be increasingly important for science. I take to heart Donna Haraway’s point that there is no way to live and also keep our hands clean.2 Instead, I want to make two points. The first is the reality of how science reform got so popular. John Arnold dumped tens of millions of dollars on a few academics with heterodox views, which catapulted them to fame and influence. When and why powerful and wealthy outsiders chose to pluck scholars and ideas from obscurity should be of key concern to those of us interested in the independence of science. And second, I want to suggest there are better and worse ways to organize science funding, which will always depend on political or monied elites. We can’t have clean hands, but we can be honest with ourselves and deliberate with our choices. **Arnold’s** **influence** is undeniable. (Sorry, Reviewer A.) Two of the main academic figureheads for science reform, Brian Nosek and John Ioannidis, both give interviews saying their ideas had little traction until Arnold, the ‘Medici of meta-research’ as Ioannidis calls him, began funding them. Arnold gave money to Nosek to create the Center for Open Science (COS) and to Ioannidis to build the Meta-Research Innovation Center. He also funded Retraction Watch and other science reform projects that popularized the replication ‘crisis’ framing and cast doubt on the trustworthiness of science. This was not merely seed money. Arnold remained the single largest funder of COS every year until 2020. Even in 2022, the last year where COS finances were public3, The Arnold Foundation was contributing at least $1M annually, and their $23M in lifetime contributions accounted for 47% of all revenue COS had ever brought in. Arnold gave a lot of money. Wired reported more than $80M to “science critics and reformers” from 2012-2017. Stuart Buck, formerly the director of research at the Arnold Foundation (later Arnold Ventures), recounts allocating “over $60 million” of that. All of that money buys a lot of things. It pays for big, expensive replication studies that appeal to vanity journals like _Science_ and bring attention to the cause. It also buys things that feel more ‘unseemly,’ such as paying the first 1,000 scientists who would use their preregistration platform $1,000 each. Buck, a self-styled “metascience venture capitalist,” further takes credit for the money and his “active particip[ation]” in keeping organizations from going under, lobbying the federal government, instituting data and journal standards, and more. In the same substack post, Buck also takes credit for the scope of the vision: the foundation cold-called Nosek in 2012. Nosek didn’t seek them out, and initially his goal “was to get us to fund a post-doc position at maybe $50k a year.” A multi-million dollar center was the foundation’s idea. It’s hard to fathom a scholar turning down such an offer. **What motivates these investments?** Why did a hedge fund guy’s foundation take all this initiative to reach out to scholars and offer them immense resources they had not asked for? And then sustain them with tens of millions of dollars over the next decade? For insight into the Arnold Foundation’s motives, I’m relying on Wired’s interviews and reporting and Buck’s own retrospective account. They begin with Gary Taubes, a journalist with a masters in aerospace engineering and a best selling book claiming that the scientific consensus and government guidance about dietary fat was wrong. The foundation reached out to Taubes and gave him $40M in the first two years to found the Nutrition Science Initiative (NuSI). In Wired’s telling, Arnold heard Taubes on a libertarian podcast and reached out to offer him the resources needed to overturn scientific consensus. They paid a journalist to dig into the review process that led to federal dietary guidelines. That’s a good way to cast doubt on any finding, but not a particularly promising tactic for answering the substantive question about diets and health outcomes. (I’ll come back to NuSI in a bit.) After his positive engagement with Taubes, Arnold said he was convinced he could no longer trust any science unless he personally knew the researcher or vetted the work–he feared “the whole foundation of the research had been flawed.” And so Arnold began reaching out to other people who said science was untrustworthy and giving them money. The foundation cold-called and funded Ben Goldacre for his claims that “pharmaceutical companies, by refusing to reveal all their data, have essentially deceived the public into paying for worthless therapies.” Nosek got a call because of a _Chronicle_ article discussing his Reproducibility Project – then a volunteer crowdsource affair that the _Chronicle_ summarized as “We want to see how much of what gets published turns out to be bunk.” The article lionises Nosek as a brave challenger of the scientific establishment, opposed by internal forces who want to protect their own reputation and that of their field. Of course, “organized skepticism” is one of the four Mertonian norms of science. We’re supposed to scrutinize scientific work. And heterodox outsiders with views that challenge the scientific establishment are sometimes right (germs do cause disease, tectonic plates move, snake venom is a chemical substance, AIDS activists were right to demand changes in research, and so on). But there is something particularly dangerous to science, and particularly appealing to billionaires, about sweeping claims like Ioannidis’ that “most published findings are false.” They cast doubt not on a substantive fact about the world, which might be revised with better science, but on the authority of the scientific enterprise itself. Arnold made fortunes at Enron and then his own hedge fund, retiring at 35. As Wired tells it, Arnold believes traders have incentives to be correct: they make money when their bets are right. For successful traders like himself, their wealth is proof that they are intelligent and correct. The notion that ‘I’m rich, so I must be the smartest’ is a sort of secular prosperity gospel. Arnold worries that scientists have career and financial incentives to publish, which is not necessarily the same as being correct. So he wants to remake science with ‘better’ incentives (i.e. in his own image). (Most scholars I know want to be correct not for the sake of our careers, but because we care about the things we work on, whether that’s finding a cure for cancer or understanding pathways out of poverty. But passion, altruism, and curiosity are hard to frame in the language of ‘incentives.’) **A lot of good** has come from one billionaire deciding to make science reform his pet cause. Some fields prone to especially small and WEIRD samples are thinking more about sampling. And there has been a huge growth in open access publishing, data, and code (although my paper with Molly shows that growth is uneven, and openness is not always good or possible). But **there is a lot of danger** , too, in allowing the whims of the world’s wealthiest or most politically powerful individuals to cherry pick scholars or ideas out of relative obscurity and catapult them into major movements.4 We wrote initially that the funding model was open to actors with worse views than Arnold. Peter Thiel also believes science is “very, very diseased,” too hostile to heterodox views, and devoid of finance industry style incentives. It would not be hard for him to spend tens of millions of dollars on his pet science reforms (ending ‘wokeness’ in his case). The reviewer did not see the relevance–Thiel wasn’t funding science reform at the time. Now, of course, Thiel does not need to. The Trump administration is using federal funding to push that agenda, with Thiel’s former chief of staff as the new head of the Office of Science and Technology Policy. Cooptation of science reform for dubious ends is not new or impossible to anticipate. The tobacco industry leveraged the virtues of open science to discredit research on lung cancer (they argued the research was untrustworthy, because the medical records used as data were confidential). For many of the world’s political and economic elite, science represents an outside authority challenging their own interests (consider climate change, vaccines, _& c_.). It is easy to imagine why some of them would want to promote scientists saying science is untrustworthy. **Where are we now?** Stuart Buck, for his part, is now executive director of a related initiative and writes at length for a Trump-linked think tank in support of the “make American healthy again” agenda, claiming “NIH needs urgent reform.” He periodically gets pulled into social media fights over whether science on vaccines and climate change can be trusted. (He also opposed some Trump admin cuts.) NuSI did end up funding scientific research testing Taubes’ theory. Taubes got into a public fight over whether he tried to censor the study, which eventually published results showing Taubes was wrong. To their credit, the Arnold Foundation pulled out of NuSI after that. This is the primary risk of people outside science cherrypicking ideas or projects to fund that they find appealing. They can be wrong. Indeed, people with no expertise _will_ be wrong most of the time when they think they know better than a consensus of experts. Most heterodox ‘free thinkers’ are wrong or worse (reactionaries, quacks, grifters…). For his part, Taubes was invited to talk about his theory in an “NIH scientific freedom lecture” last month. Meanwhile, the scientist who discredited his theory recently quit the NIH, citing censorship. Scientists receiving this money don’t necessarily share or even know about their funders’ views. Brian Nosek and several other science reform figureheads have publicly criticized some aspects of the Trump administration’s science policy. In my own little corner of academia, most of the science reform advocates are strongly opposed to these politics. Our reviewer, who seemed very embedded in open science, did not know the vast scope of Arnold’s funding, which was several times more than the federal sources they thought made up the majority of COS’s total revenue. The editors suggested that the funding politics probably did not influence scholars’ behavior because “very few people are likely aware.” That seems right to me. It’s why I’m writing this. **A better way** is possible. It isn’t even particularly new. Philanthropists funding science can empanel a group of domain experts to solicit, review, and accept proposals. Domain experts are better positioned to know what work is promising, and to understand what the contribution of new research would be. Chemists should review chemistry proposals and so on. This is how the NSF, NIH, and other public funding agencies have long worked. It’s also how the Gates Foundation, Russel Sage, the Templeton Foundation, Schmidt Sciences, and others allocate much of their research support. It’s not perfect. I wish the economists Schmidt relies on were less narrowly focused on experimental designs, for example. But it is a strong compromise between funders’ interests and scholarly autonomy. Funders under this model still set topics and priorities–a disease they care about, a place that matters to them. And scientists are still asked to justify their importance to the public or the benefactor. Funders know what matters to them. But it is a tragic flaw of hubris to think they will pick the specific scholars or answers to scientific questions correctly. A panel of experts choosing the projects and people to fund insulates the whole system against several major risks. It prevents con artists and quacks from taking advantage of funder’s relative ignorance of the science. It prevents science as a whole from lurching wildly to chase whatever whim or fad catches the fancy of billionaires and politicians.5 And most importantly, it maintains faith in science as an enterprise, where organized skepticism is incorporated as an internal part of the truth-seeking process. Peer and editorial reviews are _brutal_ , at least in the sciences that still do them.6 I don’t have a solution to the cooptation of science reform by anti-science actors. More sober language than “crisis” from the reformers would be good, but it would not prevent motivated actions like we have seen in the last few years. The language of ‘crises’ and brave ‘heterodox’ outsiders challenging consensus is an effective way to draw in some philanthropists and politicians. But as any good metascientist should recognise: there are incentives here. To attract ‘metascience venture capitalist’ money or excite the current federal science leadership, it pays to make bold, sensational claims and discredit past work. Just as publishing is separate from getting things right, appealing to the desires and egos of economic and political elites is not necessarily aligned with doing careful, correct work of one’s own. 1. During the back and forth, we were asked to add discussion of the journal’s financial relationship with Sage for the sake of balance. I was happy to add it. An editor disliked having a discussion of how the professional organization Sociologists for Women in Society depends on revenue from Sage, but allowed it to remain, unlike the discussion of COS relying on money from Arnold. ↩︎ 2. Page 236 of _Manifestly Haraway_ has a compelling discussion of how it is impossible to be innocent and we should instead aim to be honest and deliberate about what we’re doing. She’s discussing ‘invasive species’ and killing animals in that passage, but the impossibility of innocent living is a broader theme of her work going back at least to the Cyborg Manifesto. ↩︎ 3. The finances were public at this OSF link, Nosek’s account. After we cited it to correct the reviewer, the document was taken offline and the link redirected to “forbidden.” The link to the Wired UK article we relied on also broke at this time. The coincidence was alarming. So I emailed Nosek, who sent back a friendly reply and attached a copy of the missing finance spreadsheets for me the next day. He also gave me permission to cite it as a personal communication from him. Cheers, Brian! It turns out that Wired did not take the story down, either, they just changed their URL structure from wired.com/article/john-arnold-bad-science to wired.com/story/john-arnold-bad-science. No conspiracies here, fortunately. ↩︎ 4. Most infamously, Jeffrey Epstein was well known for picking specific scholars and ideas to fund, often ones that fed into his own eugenic desires. I do not know of any connections between him and the other folks I write about here. This is an extreme case of the risk inherent to a setup where funders cherry pick the things that appeal to them most and shower them with money. ↩︎ 5. Again, Epstein comes to mind. ↩︎ 6. Most writing on peer review these days comes out of computer science, where conferences get 21,000 paper submissions and a majority of authors contribute to a free rider problem by shirking review duties. It is little wonder that AI ‘reviews’ are growing in popularity there. ↩︎ ### Share this: * Share on X (Opens in new window) X * Share on Facebook (Opens in new window) Facebook * Like Loading... ### _Related_ ## Author: Jeffrey Lockhart Jeff is starting as an assistant professor of sociology at UC Berkeley in Fall 2025. He skeets at jwlockhart.bsky.social . View all posts by Jeffrey Lockhart