zeynep tufekci

My Princeton AI and society team is at the AI Safety summit in Paris. My Neurips keynote for our approach: slideslive.com/39055698/are... Our paper: ai.sociology.princeton.edu Say hi especially if you are *utterly* unimpressed by the current doom or hype models. Wrong nightmares!

This is playing with fire. It’s also a plot worth of Dr. Evil, to spark the next deadly pandemic because we can’t shut down or regulate the dwindling number of mink fur farms. Fur farms are cruel, dangerous and carry actual catastrophic risk.

"Many scientific fields are already drowning in plausible-seeming but unverified papers": this is publish or perish, not LLM. Science has been drowning before chatGPT and is just drowning more. If LLMs are used for more papers, they won't improve science. But they could be used for better papers.

"Many scientific fields are already drowning in plausible-seeming but unverified papers generated with significant help from a large language model. What’s slop and what’s not? Is there an actual gem amid the gazillion new hypotheses? It’s becoming increasingly impossible to tell."

Yes! At least from the non-medical view.

Polls show people expect AI to at least advanced science. Navier-Stokes, right? AI may throttle science instead! It doesn’t always advance understanding, its glut overwhelms filters for quality, and its methods don’t fit medical research. Gift link to my NYT essay. www.nytimes.com/2026/09/22/o...

“A 2024 review counted some 300 cases of lab workers becoming infected by the pathogens they were studying from 2000 to 2021, many of them in laboratories a step or two below the maximum-containment facilities on this map...the authors called that count “the tip of the iceberg.””

Shame on the cabal of top scientists who mocked me and other dissidents for challenging their sordid Covid origins cover-up. It is now beyond doubt that they betrayed both humanity and science - and there should be retribution. My MoS column mol.im/a/16005037

Even running a HEPA unit indoors when the AQI is in the Hazardous range outside may not improve your indoor AQI to the "Good" range. My colleague @wisconsam.bsky.social improved their AQI from Purple (borderline Very Unhealthy) to Orange (Unhealthy for Sensitive Groups) by running HEPA in the room.

Zeynep put it well

The FIFA VAR interesting as window into how algorithmic governance can operate in practice outside of the China context: highly-surveillant and high-tech (ball sensor, graphs, etc.) accompanied by a form of transparency (replays, etc.) without accountability and still essentially a blackbox.

Happy birthday, USA! 🇺🇸 Thousands of locals and tourists alike flooded lower Manhattan to catch a glimpse of the Tall Ships parade while military planes flew overtop. 📸: Angela Weiss and Joseph Prezioso via AFP via Getty Images; Vincent Alban and Jeremy Weine via Getty Images Read: bit.ly/4fmAgd5

Tall ships and airplanes along the Hudson River as they participate in the Sail4th250 in New York City, Saturday morning.

New York To Host Historic Tall Ship Parade To Celebrate America's 250th Anniversary - News 12 Brooklyn, Sail4th 250 via Marine Insight

There's a great deal that's alarming about the Prairieland sentencing, regardless of how one views the shooter, but the biggest deal is 30 years for a man who wasn't there, convicted of moving a box of zines while knowing protesters. Consider the implications. theintercept.com/2026/06/23/p...

I’m honored to be giving the 2026 US-UK Fulbright Distinguished Lecture, titled “Are We Having the Wrong Nightmares about AI?” The in-person event in London is completely full, but there is a free livestream: www.eventbrite.co.uk/e/dr-zeynep-...

The New York Times is hiring a national science correspondent job-boards.greenhouse.io/thenewyorkti...

NEW: FCC to direct eight Disney-owned TV stations to file early license renewals, a source says, adding that the "unprecedented" action was directly tied to ABC late-night host Jimmy Kimmel's recent joke about first lady Melania Trump.

Tufekci's keynote here remains the best framing for how to think about AI, IMO. The key changes will be in what things become cheap/easy and what things remain expensive/hard. AI will remove a lot of friction. Some of that friction is load bearing. bsky.app/profile/zey....

“A whole civilization will die tonight,” Trump declares.

I reviewed @mollycrabapple.bsky.social 's terrific new history of the Bund — a great read but more importantly, a book that could not have arrived at a more urgent moment. www.nytimes.com/2026/04/06/b...

The AP is reporting that Sharif University in Tehran has been bombed. This is Maryam Mirzakhani’s alma mater — the first woman to win a Fields Medal in math, and only one of two so far. The university is also a leading center of opposition to the government. apnews.com/video/ict-bu...

Some science cuts in Trump's proposed FY27 budget ❌ NIH: -$5B, including all of NIEHS, NIMHD, FIC & NCCIH. ❌ EPA: -4.6B ❌ NSF: -3.8B ❌ NASA Science: -3.4 Billion ❌ NOAA: -1.8B, including the entire Oceanic and Atmospheric Research budget ($605M) ❌ AHRQ: All of it. -$296M, 100% And there's more.

The article makes this fairly explicit, but it's worth emphasizing that Grammarly doesn't have any clear option to request a specific reviewer (hence you can't easily check if you're included) and even a cursory attempt gets it to produce a bunch of random scholars from my or adjacent fields.

Well, there it is.

"This firm’s special treatment is a stark example of the way that the second Trump administration has used taxpayer money to benefit people close to the president or his top officials." www.nytimes.com/2026/03/18/u...

Trump's DOJ this month quietly backed Alexander Smirnov's bid to withdraw his guilty plea & wipe out his sentence. Smirnov had admitted lying to the @FBI about Burisma's owner arranging to pay $5m in bribes to both Joe & Hunter Biden. 👀 @davidcorn.bsky.social catch:

The State Department is considering withholding lifesaving assistance to people with HIV in Zambia as a negotiating tactic to force the government of the southern African country to sign a deal giving the United States more access to the country’s critical minerals. www.nytimes.com/2026/03/16/h...

I was impersonated by Grammarly as well. Besides the lack of consent, a big problem is that the LLMs do a highly plausible impersonation of my views for anything not specific to me but miss exactly where my views differ from many AI experts or critics. So worse than useless. Actively misleading.

PSA: One consequence of AI language models is that you should check your spam folder regularly. LLMs can generate targeted/tailored outputs at scale, resulting in a growing overlap between spam and email meant for you. Filters are failing: more spam in your inbox AND more of your mail in spam.

When a federal judge (finally) put ICE officers under oath, they admitted that they are given daily detention quotas and rely heavily on a Palantir-supplied AI tool to select targets, without warrants and without enough evidence to obtain one. They simply go into neighborhoods and round people up.

Five-byline alert: 🚨 “.. An ongoing military investigation has determined that the United States is responsible for a deadly Tomahawk missile strike on an Iranian elementary school, according to U.S. officials and others familiar with the preliminary findings.” www.nytimes.com/2026/03/11/u...

Whistleblower complaint says the DOGE engineer "told another colleague, who refused to help him upload the data because of legal concerns, that he expected to receive a presidential pardon if his actions were deemed to be illegal." www.washingtonpost.com/politics/202...