Sara Goudarzi

AGI gets thrown around a lot. So, for the @thebulletin.org I spoke to researchers and took a look at the term and reality of AI that could rival human intelligence. Turns out it's a better tool for raising capital than a scientific milestone.

U.S.-China cooperation on AI shouldn't wait for an agreement on model restrictions. In @thebulletin.org, I propose a joint 6-month pilot using simulated data to test AI applications for preventing nuclear theft, sabotage, and terrorism. thebulletin.org/2026/10/how-...

Anthropic logged net loss of over $80 billion in first half of 2026, WSJ reports reut.rs/4z2Zi9J

My latest for @thebulletin.org. You might be wondering, why are we still talking about the beleaguered Climate Working Group report a year+ later? Because Florida's DOGE and others can still cite it in reports about “squandering millions in virtue-signaling about a so-called climate crisis.”

🤔

AI doomerism isn't sound science, it's a parlor game. Great to see reporting that pushes back and quotes experts like @timnitgebru.blacksky.app and @emilymbender.bsky.social. Coxon and Anthropic "did not respond to a request for comment." 👀 👀 👀 @thebulletin.org thebulletin.org/2026/09/how-...

One of the handful of articles telling it like it is instead of misinforming the public. By @saragoud.bsky.social for @thebulletin.org Yes. Its bad science based on numbers out of various behinds, but that's not how most of the rest of the media is discussing it. thebulletin.org/2026/09/how-...

Anne Carson.

My toaster went rogue this morning when I left it on air fry, at 400 degrees, for 10 minutes, and was shocked to find that it encouraged thousands of highly coordinated air molecules to move rapidly, completely burning my bagel to a crisp. I only wanted it lightly toasted

These probabilities make the science look settled and collapse layers of uncertainty, explains Sarah Kreps. "Experts still cannot establish with confidence what AI models can do, the factors that might limit the current trajectory, or the threat models underlying catastrophic harm."

Anti-Prime Day starts…NOW! 🥳 Today through midnight on 10/7, enjoy free shipping site-wide, all while supporting local, independent bookstores.

Calling all NYC artists — painters, filmmakers, poets, musicians and more! Did you know our city provides grants to support your work every year? And this year, we're DOUBLING the amount — from $5,000 up to $10,000 per artist! Deadlines are approaching fast — apply today at nyc.gov/artistgrants.

thebulletin.org/2026/09/how-...

Friday’s Page One. @financialtimes.com

“There is no established path showing that today’s #AI systems will surpass the smartest #humans, much less a #scientific basis for calculating the likelihood that such #systems will destroy #humanity.” #AIdoomerism thebulletin.org/2026/09/how-...

“What we have to keep in mind is that all this doomerism isn’t based on any scientific work,” says Boyan Milanov, Senior Research Scientist at AI Now Institute. “It really is science fiction. None of the claims are scientifically verifiable or falsifiable.” thebulletin.org/2026/09/how-...

AI companies are calling for regulation in ways that let them set the rules, not independent parties who can hold them to account. “If you read between the lines, everything [Amodei] advocates [for] is something that would help companies like Anthropic,” says AI Now's Boyan Milanov bit.ly/3TxmJbq

AI doomer discourse has been in overdrive the last several weeks: a resignation that set off a media frenzy, endless warnings, and even a Senate bill to ban superintelligence. But where’s the science and why all this hype? I dug into how doomerism serves big business.

There's an issue behind warnings about superintelligence, writes Sara Goudarzi @saragoud.bsky.social. "There is no established path showing that today's AI systems will surpass the smartest humans, much less a scientific basis for calculating the likelihood that such systems will destroy humanity."

AI doomer discourse has been in overdrive the last several weeks: a resignation that set off a media frenzy, endless warnings, and even a Senate bill to ban superintelligence. But where’s the science and why all this hype? I dug into how doomerism serves big business.

United States President Donald Trump issues executive order instructing federal government agencies to "use the terms 'Super Intelligence' and 'SI' in place of 'Artificial Intelligence' and 'AI' and will not acknowledge the usage of 'Artificial Intelligence' and 'AI' in any applicable setting."

In case you missed it: "Insiders are betting on war: How prediction markets are becoming a national security threat."

"Do not fret. If you absolutely cannot make that 8:30 a.m. meeting, we will still have it without you. And we will have it again when you are able." buff.ly/k2hhGaM

i'm tired of seeing “escalating ai capabilities” reported as if it is a fact. let's be clear this claim does not rest on transparent, well-defined, independently validated metrics, but is inferred from model behaviours elicited by humans and measured through opaque, poorly designed evaluations

Anyone with insider information can place bets on prediction markets, creating an operational security nightmare by giving allies and adversaries advance warning, write David Szakonyi, Michelle Kendler-Kretsch, Michael Hornsby of the @acdatacollective.bsky.social

"Prediction markets are creating new nightmares for operational security," write @szakonyi.bsky.social, Michelle Kendler-Kretsch, and Michael Hornsby. They allow those with inside information to bet on national security issues, all while giving both allies and adversaries a possible heads-up.

Anyone with insider information can place bets on prediction markets, creating an operational security nightmare by giving allies and adversaries advance warning, write David Szakonyi, Michelle Kendler-Kretsch, Michael Hornsby of the @acdatacollective.bsky.social

Amid the tsunami of headlines claiming that AI agents went “rogue” in the Hugging Face breach, @eryk.bsky.social offers a much-needed analysis of what actually happened, and why the incident was the result of human decisions and not autonomous AI behavior.

ahem exactlyyy

At the core of the Hugging Face hacking incident were a series of human choices that traded security for speed, writes Eryk Salvaggio @eryk.bsky.social. Focusing on the human decisions allows for a more rigorous analysis of what went wrong and what can be done about it.

Amid the tsunami of headlines claiming that AI agents went “rogue” in the Hugging Face breach, @eryk.bsky.social offers a much-needed analysis of what actually happened, and why the incident was the result of human decisions and not autonomous AI behavior.

"Frontier laboratories are building architecture that is centralized three times over: The same firm trains the model, owns the infrastructure it runs on, and revises unilaterally the rules for who may speak with it and on what terms," writes @allisonstanger.bsky.social for @thebulletin.org.

"Frontier laboratories are building architecture that is centralized three times over: The same firm trains the model, owns the infrastructure it runs on, and revises unilaterally the rules for who may speak with it and on what terms," writes @allisonstanger.bsky.social for @thebulletin.org.

Rough Draft vs Final Draft

New excerpt from my forthcoming book Bytes and Bullets in @thebulletin.org. Take a look!

"Any major armed conflict fought today will almost certainly involve drone technology. Consequently, UAVs are changing the character of war and challenging assumptions about how military victory is achieved," excerpt from @stevefeldstein.bsky.social's “Bytes and Bullets” in @thebulletin.org

“It is clear that we should expect to continue to experience such catastrophic events as these in coming years. Our climate is set to warm significantly in coming decades, which will only continue to contribute to further destabilisation of such high mountain environments.”

NEW @thebulletin.org by Neta Crawford: The US military's fatal stance on climate change

Officials who've dealt with the complexities of data center deals are largely unwilling to talk about them due to public backlash. The result is a self-defeating information gap: local governments negotiating their first deals lack access to experienced people to ask for advice, writes Abi Olvera.

Local officials who’ve navigated the complexities of data center deals aren’t sharing notes, fearing backlash. The result? A self-defeating information gap that makes it harder for communities to learn from one another, writes Abi Olvera @thebulletin.org's disruptive technologies editorial fellow.

Two people have died in Pennsylvania from measles, the state health department said. Both were unvaccinated. These the first deaths related to measles within the state in 35 years, it reports

A new interactive tool from the Bulletin illustrates the connections congressional officials have to defense contractors and military tech firms. Check it out here: buff.ly/WUtWi5r

My last piece as editor in chief for the @thebulletin.org is, I think, (and thanks mostly to the brilliant @thomas.gaulk.in) beautiful and powerful: thebulletin.org/2026/08/the-...

Can AI help someone build a lethal drone? Matt Smith documented his experience finding out. "The creation of an autonomous weapon with equipment easily available to the public seems possible, but the process of doing so was—for me, at least—more difficult and stranger than expected."