Carl T. Bergstrom

Carl T. Bergstrom

@carlbergstrom.com

Professor, UW Biology / Santa Fe Institute I study how information flows in biology, science, and society. Book: *Calling Bullshit*, http://tinyurl.com/fdcuvd7b LLM course: https://thebullshitmachines.com Corvids: https://tinyurl.com/mr2n5ymk he/him

150.2K followers2.6K following13.5K posts

Pinned

I'm thrilled to share my first foray into nature writing. I describe my travels through Aotearoa New Zealand to see the country's rarest birds—and use this as a vehicle to tell the stories of the islands' evolutionary history and the bold conservation efforts that saved these creatures. Gift link

bloomberg.comThe Birds I Waited 45 Years to SeeAn evolutionary biologist returns to New Zealand to find the rare birds he dreamed of seeing as a child making an extraordinary comeback.

What kind of numerals do you use when you make up fake statistics about vaccine harms, you shitweasel?

The Bulwark@thebulwark.com

RFK: "There's five hundred million Arabs in the world. And yet they don't produce science. They don't win the Nobel Prize. They don't win the Fields Prize. They don't win the Wolf Prize. They don't produce literature."

"AI can draft an essay for you, especially if you give it a high quality prompt." Really? This is what MIT thinks high schoolers should learn? MIT's "Day of AI" dayofai.org provides materials for K-12 teachers and claims to have reached over 5,000,000 students to date. 1/2

Your monthly reminder that if you're using ChatGPT to do your science, you're failing your colleagues, community, and all of us. Here, ChatGPT makes up data and presents it as the real thing — then apologizes when confronted.

View full thread

In my opinion, it’s completely beside the point because people who need this warning will use all sorts of models, free and paid—but because the LLM apologists are screaming, this is OpenAI’s latest subscription model, Astra 6, prompted to fetch and plot the original data.

Your monthly reminder that if you're using ChatGPT to do your science, you're failing your colleagues, community, and all of us. Here, ChatGPT makes up data and presents it as the real thing — then apologizes when confronted.

Also for what it is worth, it DIDN'T actually label them as illustrative — and the deception was only apparent because I know the data well. That's scary, because if you use AI for everything, you wouldn't know the data well enough to catch its deceptions.

On one hand, as guy living in the US I acknowledge that that I have very little right to comment on political matters in Brazil. On the other hand, if you're coming at me from the *left* upset that I'm happy Lula won? What the fuck team are you actually on?

Carl T. Bergstrom reposted

Wait wait! In addition to all this: Real Clear isn’t some spontaneous thing, it’s a rightwing site funded by the same far-right billionaires - Koch, Uihlein, Bradley, etc - that want to tear down the govt, incl to destroy US cancer research. Real Clear is part of the Project 2025 effort.

Carl T. Bergstrom@carlbergstrom.com

So in a huge surprise to absolutely no one, it turns out that NIH director Jay Bhattacharya's new journal based on the premise of letting his friends and only his friends publish essentially whatever they want is full of crackpot nonsense from retired cranks.

Let me propose that they rebrand as the "Journal of Applied Agnotology".

Carl T. Bergstrom@carlbergstrom.com

So in a huge surprise to absolutely no one, it turns out that NIH director Jay Bhattacharya's new journal based on the premise of letting his friends and only his friends publish essentially whatever they want is full of crackpot nonsense from retired cranks.

A bit earlier, but I'm old enough to remember when Ann Arbor got its first Mexican restaurant.

Peter@notalawyer.bsky.social

a 25 year old could not comprehend how bad the dining options were in the median American town in 1996

Sure but that's also what the over-leveraged oligarchs would want you to think. Someone has to go down with the ship.

The Washington Post@washingtonpost.com

You don't need to check on your 401(k) daily — and doing so may backfire. “Frequent monitoring can increase anxiety and tempt investors to react to normal market movements rather than stay focused on their goals,” one financial adviser said.

So in a huge surprise to absolutely no one, it turns out that NIH director Jay Bhattacharya's new journal based on the premise of letting his friends and only his friends publish essentially whatever they want is full of crackpot nonsense from retired cranks.

publichealth.realclearjournals.orgAre air quality‒health effect claims a statistical analysis fluke?There are literally thousands of epidemiology (observational) studies in the literature claiming that air quality causes adverse health effects, including mortality. There are three profound examples ...
View full thread

What's so striking about the original piece is how utterly incoherent it is. It's logical consequence of pursuing the "peer review is censorship" mentality that Jay has espoused ever since everyone and their undergrad RA pointed out that his April 2020 study was hopelessly flawed.

So in a huge surprise to absolutely no one, it turns out that NIH director Jay Bhattacharya's new journal based on the premise of letting his friends and only his friends publish essentially whatever they want is full of crackpot nonsense from retired cranks.

publichealth.realclearjournals.orgAre air quality‒health effect claims a statistical analysis fluke?There are literally thousands of epidemiology (observational) studies in the literature claiming that air quality causes adverse health effects, including mortality. There are three profound examples ...

Sandhill cranes have returned to the Minnesota farm field where my father was born. His mother remembered them from her childhood, but he never saw them until a decade or two or so back.

Social media loves talking about how everything is getting worse. But what’s something that you’ve seen get _better_ in your lifetime?

talk about tech debt...

jim mallet@wtf-r-species.bsky.social

It's not just maize, it's every genome. Currently, my favourite organism, Heliconius butterflies, is one of the champions of mess -- Fang, B., & Edwards, S.V. 2025. Pangenomes: new tools for ecological and evolutionary genomics. Trends in Ecology & Evolution 41:230–244. doi.org/10.1016/j.tr...

Breaking news: students given laptops in school watch youtube.

NBC News@nbcnews.com

A horrified mother discovered her son was scrolling sports highlights and AI slop in class. NBC News' first-of-its-kind analysis shows how YouTube keeps some students hooked.

Carl T. Bergstrom reposted

This week, NIH advisors rebuked RFK for harming #NIH and US science. Sheryl Gay Stolberg reported it. Then RFK posted calling her sources anonymous or corrupt. I'll say this hoping it does good: I'm an NIH scientist, I'm not anonymous, and RFK Jr is destroying NIH. More in attachment:

WTAF!?! Watch the head. It stays looking forward.

Cell Press@cellpress.bsky.social

Ever see a snake do a barrel roll? Scientists disentangle the locomotor kinematics of this behavior in rolling snakes. Learn more in Current Biology: spkl.io/633287ntaA Yosuke Kojima & colleagues @currentbiology.bsky.social

Setting this up has been fascinating. In test sessions with GPT-6, it gets confused. It supposedly knows that I am trying to guess whether it is an AI or a human, but instead it reverses the roles and starts guessing whether I am human or an AI. Is it confused, or fucking with me?

“The culmination of my life’s work may have been targeted by an AI-assisted censorship tool & terminated by a politically-motivated AI hallucination” If you care in any way about science, you need to read this thread from my friend Kelton Minor.

Kelton Minor@keltonminor.bsky.social

Together with other scientists I’m suing the 🇺🇸 US National Institutes of Health for violating our 1st Amendment rights & censoring our health science. I’m ready to shine a light on the darkest days. Buckle up, read on. This gets pretty wild… THREAD

Carl T. Bergstrom reposted

Replying to @carlbergstrom.com

Reminds me of getting carded (in my 40s!!) at one of Seattle's fake-speakeasies-for-wankers and getting a blank stare when asking the bartender if that didn't sort of break the spell.