As @amostoh.bsky.social and I explain, the Pentagon's dispute with Anthropic should not distract from the broader crisis at hand: Congress's failure to regulate some of the riskiest uses of AI - namely, amplifying surveillance and automating the use of lethal force.
The expanding war in Iran brought to the fore questions about the role of technology in armed conflict, including the controversial use of new artificial intelligence technologies. Tech Policy Press invited perspectives from experts on what they are watching for as the situation unfolds.
Guardian headline: Israel orders more than 500,000 people to evacuate Beirut's southern suburbs
This is a war crime.
Anthropic is not the victim. They are a villain. The victims are the people being killed in an illegal war. And the victims are all of us being asked to accept a world where imprecise tools are directing lethal force.
Whether it's ChatGPT or some other company, no one should be building tech that enables killer robots & mass surveillance.
Congress should ban both of these things. Until they do people should be boycotting to hit companies like ChatGPT where it hurts! Pledge to boycott now: quitgpt.org
I am really struggling to watch the focus on Anthropic's rights as a government contractor.
Rather than on the US government using LLMs to kill people.
Dario and Sam both want technology with knowingly high error rates embedded in warfare.
WE CANNOT NORMALISE THIS.
One surprising side effect of the US war in Iran: American farmers are staring down soaring fertilizer prices at the worst possible time.
From @mollytaft.com
Subject line from The World newsletter of the NYT. "Iran chooses chaos"
Oh diiiiiiid they?
TONIGHT (5:30pm MT): Join us to explore how we can hold Big Tech accountable.
From the gig economy to data centers, we’ll discuss how we can use legal and policy mechanisms to advance people-first Big Tech accountability in Colorado and beyond. luma.com/7ey0m3ar
Starmer will go down as one of the weakest, most ineffective leaders in British history.
Altman is such a lying piece of shit.
No cookie!
The thing I have been reflecting on is that if it was any of the other guys we would never have known this moment happenened. Plus nice to have this headline even if it is motivated for all the wrong reasons.
This is from Hegseth's post. I'm no expert in this authority but am I understanding correctly this would mean that any company which has ANY contract with the U.S. military (Google, Microsoft, etc...) is now required to immediately cancel their Claude subscriptions and must divest from Anthropic?
How can we dream up a better world if we outsource all of our critical thinking to AI?
In our final episode of AI Lingo Bingo, we’re tackling some of the buzziest buzzwords with some incredible people...
And now he's at OpenAI selling ChatGPT to nation states!
Imagine what he would have done with that while still in government and what he'll be pitching to others...
"Is Claude out of the war business?" @amostoh.bsky.social was thinking hard about AI, the law, and human rights long before it was cool, and you should spare 10 minutes to listen as he joins @alixdunn.com to dig into the dispute between Anthropic and the Pentagon.
Want analysis on the Claude Venezuela situation from a legal analyst and researcher focused on the militarisation of AI?
Listen to Amos Toh (@amostoh.bsky.social) of the @brennancenter.org break down the legality and the possible commercial implications here:
www.themaybe.org/podcast/is-c...
Big Tech is lobbying our governments to expand surveillance, capture public resources for compute, and consolidate power. How can we push back with policies that put people first?
Please please please can we not do this with AI too?
Oops. Turns out Persona, Discord’s age-verification service, was secretly screening selfies against government watchlists AND accessible by the feds. It's backed by Peter Thiel.
“The state wants to see everything. The corporations want to see everything. And they've learned to work together.”
The emptiness of this man is breathtaking.
What happens when you feed both wolves ONLY PASTA for like 3 years?
Despite working 60+ hour weeks, many Uber and Lyft drivers have seen their pay drop by nearly 50% over time. Even as these companies report record profits, consumers face inflated prices, while the workers who are (quite literally) driving these companies’ success are paid less and less.
We learned today that my daughter has been turning up to school in character as a wolf.
When they ask her to do things she says that wolves don't work.
I feel like we have done something wrong, and also clearly something very right?
📻 AI Summit Lingo Bingo Pt 2 is live!
It includes insights from:
- @meredithmeredith.bsky.social
- Usha Ramanathan
- @audreyt.org
- @abeba.bsky.social
www.themaybe.org/podcast/ling...
#ComputerSaysMaybe livestream on Mar 2nd with three amazing guests fighting big tech exploitation-maxxing.
-@seligmanforag.bsky.social
-@bedoyausa.bsky.social &
-Elliott “El’Bo” Awatt
RSVP here + we'll send the recording after:
luma.com/7ey0m3ar
Please share with your networks!
“The narrative that we need big AI models—and quasi-infinite amounts of energy—tries to sell us the idea that this is the only kind of AI we need, and the only future that's possible,” says AI and sustainability researcher Sasha Luccioni. “But there are so many different, smaller and more efficient models that can be deployed for a fraction of the cost, both to people and the planet.” In a separate piece of research also published Monday, Luccioni and Yacine Jernite, head of sustainability at AI company Hugging Face, looked at the costs of training a wide variety of AI models, finding that massive proprietary models trained on access to vast amounts of data and energy aren’t the only option for powerful AI solutions. Often, smaller models perform just as well as the more expensive ones in AI application.
In the sections above, we showed that training and deployment costs across all of the categories we have surveyed — task and domain models, small and medium language models, and large open-weight and proprietary systems — vary by orders of magnitude. This report provides quantitative estimates of those cost to show that the majority of commercially and scientifically significant AI development does not require 9-figure training runs or exclusive access to data-center-scale infrastructure. Open science and open-source developers have demonstrated that models that are competitive on a wide range of benchmarks can be trained for costs that range from the low thousands to the low tens of millions of dollars, and that deployment can be as simple as running on a single GPU instance or on-device, or as scaled as paid API access. The gap between that reality and the narrative centered on frontier-scale systems is not merely technical; it has implications for who can build, audit, and govern AI.
Some coverage of our new report in @wired.com by @mollytaft.com!!!
Really want to highlight this new work from @sashamtl.bsky.social et al though: on the 'none of this has to be this bad' point -->>
www.wired.com/story/big-te...
@alixdunn.com asked @meredithmeredith.bsky.social, President of @signal.org, for her take on “open source AI” as part of our Reframing Impact: AI Summit 2026 series, in partnership with @ainowinstitute.bsky.social and Aapti Institute. Watch the full interview here: www.themaybe.org/reframing-im...
“Kenya had this health data partnership with the United States where it would provide access to the US with sensitive Kenyan health data for basically decades. That's unreasonable.”
- @chinasa.bsky.social
OKAY BIG ONE FOR YA
NEW report out today: I dig deep into how big tech has been muddling tech types and overstating evidence to claim "AI" will make global emissions drop like a stone....
Not only will it not, it seems to be verifiably doing the opposite!!
ketanjoshi.co/2026/02/17/b... + 📼+🧵>
