I am going to need every single person who said that misinformation was a moral panic to feel enough shame and regret that they leave public intellectual life.

Related point, if you are in misinformation research and have been launching study after study to see what Uncle Fred thinks after seeing misinformation PLEASE listen to what @katestarbird.bsky.social and I have been saying for years: effects on elite and official action is what matters most.

I'm with @devezer.bsky.social, you *absolutely* cannot test for p-hacking. These tests universally assume p-hacking, and nothing else under the sun, explain deviations from idealized distributions and then claim to have found it.

Fantastic thread.

For example if someone wants to tell me about space travel, I should treat that information differently if they are a friend I trust who has a PhD in astrophysics, versus some random person on the internet, versus a psychotic billionaire trying to sell a space travel company 9/

For $400k a year, I’d ask Claude how it’s doing each day and send Dario Amodei a thumbs up emoji on slack.

AI makes continuous reproducibility and robustness testing trivial. What happens to science under new levels of scrutiny and stress-testing by default? Some thoughts on how this could play out, informed by watching open science play out over the last decade.

Many are appropriately outraged by Altman’s comments here implying that raising a human child is akin to “training” an AI model. This is part of a broader pattern where AI industry leaders use language that collapses the boundary between human and machine. 🧵/

"There are deeper strata of truth…there is such a thing as poetic, ecstatic truth. It is mysterious and elusive, and can be reached only through fabrication and imagination and stylization." - Werner Herzog

My old boss and current boss happened to meet halfway across the world today, and im truly hoping they took the piss out of me.

People excited about computational reproducibility with LLMs are gonna lose their minds when the llm tells them that it’s largely a solved problem with dozens of suitable tools and it just requires knowing basic software development.

Great post worth reading. In working on a related comment earlier today, it occurred to me that we don’t discuss often enough how the fields most concerned about science (and most involved in reforming it) are the ones most wedded to NHST as the arbiter of truth.

Because of course it's an EA outlet (and presumably an EA author). They really can't help themselves, it seems. "I've written 2000 words already, can I do a little eugenics, as a treat?"

I read the piece in reference and it’s truly a tour de force of abysmal argumentation from start to finish.

Yes. One thing every scientist should probably learn is that falsification can't escape the Quine-Duhem problem. In this case, if your toy statistical model of science predicts one thing, and you see another, you can't tell whether you refuted your hypothesis or your model is wrong.

“We should be careful not to marginalise questions of power, because in wearing that aura of clean objectivity, metascience risks becoming strangely depoliticised.” By @batoolmm.bsky.social

It’s of course good to share data, but don’t take your advice on what’s safe or ethical to share from #metascience Bluesky. Legally it’s important that the source for the data allows sharing (some don’t!), and you may want to CYA with ethics approval and legal review if there’s a risk of defamation

This was one of the concerns with our paper on disclosable ties to industry. Somewhere around 200 or so authors in our dataset are indicated as tied to industry, which is a lot of opportunity for a SLAPP suit. www.science.org/content/arti...

Didn’t think being interviewed by @brandyzadrozny.bsky.social would get me into the Epstein files but here we are.

On the publication bias discourse, I regret that metascience has become a source of decontextualized, low-res, bean-counting-focused `science is in crisis' narratives. It is largely uncurious abt science, desperately lacking in theory & measurement. I'll quote a few takes I liked & add my thoughts🧵

Great piece by @cathleenogrady.bsky.social ! Featuring words from a number of level heads, including @manuelansede.bsky.social @jbakcoleman.bsky.social @richvn.bsky.social @jacksonwryan.com and others.

Also, Joe has spent even more time on this particular claim than I have, and this short thread of his makes good sense.

I have plenty of reservations about the reliability of funnel-plots and z-curves and such, as well as their interpretation..... But holy shit look at that.

A side note in metas court room drama is this little memo from 2018 discussing the goal of converting ten “academic allies” given that researchers were associating harms to their products.

Just to put a pin in it, that hyped news coverage drove moltbook to a $100 million dollar market cap, the vast majority of which has been erased in just 8 days.

It's interesting to see so many academics take the position they "only learned later" what Epstein had done. Do people not google funders? As early as August 4th, 2006, his wikipedia page included discussion of the charges, and they've been a fixture since. en.wikipedia.org/w/index.php?...

"Steven hawking was in a wheelchair and we thought he was pretty smart" is a wild argument to make in a scientific article. www.nature.com/articles/d41...

There's something truly absurd about a paper blaming science itself for gaps in trust, led by an author whose work was retracted for failing to follow the exact recommendations in the piece. www.pnas.org/doi/10.1073/...

Comment dit-on “get his ass” en Francais? www.bbc.com/news/article...

It’s somehow a breath of fresh air to see the obvious stated… peer review requires peers!

New draft: "Decline effects, statistical artifacts, and a meta-analytic paradox". In this manuscript I show how a common practice in meta-analysis (eg the 2015 Open Science Collaboration) creates artifactual signatures of poor scientific behavior. PDF: raw.githubusercontent.com/richarddmore... 1/x

🧪 It's been 10 years since Dorothy Bishop and I published a commentary in Nature about the risks of transparency. doi.org/10.1038/529459a 1/10

Tech journalists try googling the founder or talking to literally anyone with a clue before writing a breathless story about the latest bullshit challenge.

The guy who vibe coded a crypto rug pull scientific error detection and got coverage in Nature is back with a vibe coded ai social media network, with similarly crypto intentions… getting coverage from major outlets. www.404media.co/exposed-molt...

Another mess from Bhattacharya at "Reclaiming Science"... He talked about John Ioannidis's paper "in 2006"... "with the audacious title 'Why most published, peer reviewed scientific research papers are false'" that is "three pages long". 1/9

With the deluge of LLM science, it's sooner or later that we have to move to a model of evaluating scientists for promotion and funding on a finite number of papers per year.

On the bright side, latex generated this way will probably be easier to filter from places like @arxiv.bsky.social

Fantastic morning read. "This leads us to ask whether delegating judgment to LLMs ...transforms it into what we call epistemia... producing the illusion of knowledge when surface plausibility substitutes for evidence-based reasoning." www.pnas.org/doi/full/10....

Home vibe science is killing paper mills. www.technologyreview.com/2026/01/27/1...

These latex errors feel enough like human generated science to convince me LLMs are sentient.

I haven't played with this much, but it looks like sort of a "one-quick trick" to ensure computational reproducibility. We could probably skip a couple of decades of science reform re: reproducibility if we could get broader uptake of existing tools like Lix/Nix/etc... lix.systems

How many LLMs can be persuasive paper are gonna be published in top journals before we remember that yes, people can be persuaded by text? And where was this thinking amidst all the hot takes that social media and misinfo has no causal effects?

"The reason these aspects of our jobs have been so challenging to automate is that they rely on something even more precious than our time: namely, our capacity for scientific decision-making. It is worth considering what we lose when we cede that—and our agency—to machines."

This is a good step, and I’ve always felt these endorsement mechanisms provide an interesting opportunity for networked sanctions—e.g. you ban the account, suspend their endorser for a bit and withdrawal their endorsement capacity.

This is a great take. His video discussing how the sanitation trucks become snowplows and we put chains on bus tires felt like a factoid from an off duty tour guide, who subsequently goes on about converted firehouses.

Perhaps the most politically important thing about Zohran Mamdani is that it seems like he really, really likes New York and New Yorkers.

A reminder that we can expect and must demand so more from political leaders—including moral clarity. I was fortunate to have Ami, a great speechwriter, work with me on several White House speeches. Read what he would have Schumer say in an alternate parallel universe of courage and conviction 🧵

The @nytimes.com got the titles here much better than usual.

Truly stunning leaf out of my frydek.