“We should regulate AI” is such a meaningless statement. How? Say what you actually want.

There’s no magic regulation fairy that makes AI safe.

In 2018, Meredith Whittaker organised a walkout at Google, in protest at an AI defence programme called "Project Maven".

Whittaker says this project - dropped by Google but taken up by a rival firm - was responsible for a 2026 US airstrike on a school in Iran which killed 150 y hasn't yet released its Minab strike findings. The US Defence Secretary says they don't intentionally target civilians.

🎧 How did "AI delusion" impact the lives of vulnerable people around the globe? bbc.in/4nx49uB

@mer__edith on @BBCNews - "Question why we are mediated by seven platforms largely jurisdictioned in the US that make us supplicate to a format that degrades the integrity of our expertise." 🔥

Meredith Whittaker is the president of Signal, one of the world’s most prominent encrypted messaging apps. She has also been called one of the most powerful women in AI, and has ridden through the ranks within the tech world.

🎧 bbc.in/4rYI4rl

@Sukkola ”On hyvin kiehtovaa ajatella, että talossa asuu hirviö, vajassa kummitus tai että meitä uhkaa monumentaalinen vaara. Tuon tunteen vuoksi katsomme elokuvia ja käymme vuoristoradoissa. Nyt tarvitaan kuitenkin viileitä, harkittuja ja aikuismaisia näkökulmia.”

- Meredith Whittaker

“If I do nothing else in the world, I want to defend art as a deadly serious existential requirement for human flourishing & the possibility of a liveable future…That’s a belief core to my being, & I see the consequences of the disavowal of art all around me.”

Me? With my claws and teeth I will will defend art and artists as essential to any version of a beautiful livable world.

a-rabbitsfoot.com/editorial/conf…

Oh, hi!

No, "anthropomorphizing" is a technical problem, about how AI works, not something like woke language policing. We say it's wrong because it's "technically" wrong not "morally" wrong.

If you fail to set the brakes when parking on a hill, and your car then rolls down the hill, you don't claim that the car "decided to drive away on its own". Such anthropomorphizing distorts what really happened -- that you failed to set the brakes.

The same applies to AI. When you "anthropomorphize" the AI you fail at understanding what actually happened.

Registration for The People's AI Assembly is live!

On Oct 19 with @MorePerfectUS, we’re bringing together leading organizers, cultural figures, and policy leaders shaping the frontlines of AI. It’s time to build a tech movement for us!

Join us in NYC: bit.ly/ai-assembly

These are precisely the scenarios that I spend a significant time actually being concerned with. The adoption of unreliable, brittle, and inaccurate LLMs within safety-critical fields (e.g., defense and nuclear) is not only a concrete risk, but has catastrophic consequences.

AI Safety & Industry Thinking 📖: a personal anecdote

The recent cyber-security screw-ups with AI-mediated attacks on third parties reminded me of a discussion a few years ago...
Story time! 👇

La rentrée ce dimanche du @Grand_Continent avec des professeurs qu’on aurait aimé avoir—@Tyler_A_Harper, Haenel, @CareniniE, Jankélévitch—Google et le miso de patate douce par @mer__edith, les enfants de Bruegel, @antoniospadaro sur JD Vance et le gouvernement de l'apocalypse.

This was always going to happen to age verification systems. It wasn’t even so much of a prediction as it was an inevitability. The second half of this is tying actual web usage to real-world ID, and breaches will give us that too, soon.

@dwarkesh_sp's summary of the @OpenAI @huggingface incident has hit a nerve, but it is dangerously misleading. Sure, the @OpenAI agents did unexpectedly bad things - underlining the need to massively improve evaluation/sandboxing. But the language Dwarkesh uses is permeated by innumerable unwarranted anthropomorphisms, obscuring the lessons we should be drawing.

Examples: “from the AI’s perspective, it probably felt like that had spent a human-subjective-week of just banging their head against the wall”. No. The agents do not experience time. They do not experience anything.

“they became giddy with excitement”, “PHASEONE 10841 had discovered”, “the agents naturally assumed”, “it thought it had also been poisoned”, “the agents … desperately wanted”, “they still needed to figure out” No. Agents lines of code. They do not feel emotions, assume things, think things, want things, or figure things out.

“A lot of … agents from the second civilisation died trying”. No. Besides the hubris of the word ‘civilisation’, agents do not die because they were never alive. (The idea that agents “die” comes up multiple times in the essay.)

“On Twitter, people were debating whether the agents were truly sacrificing themselves for the swarm, or whether they were doomed anyway and so might as well try to help their peers”. Neither. Agents do what their code tells them to do, just as water finds its way down a slope. They cannot ‘truly sacrifice themselves’, since they are neither conscious nor alive.

Why does this matter? If we attribute agents with properties they do not have, then (i) we distract attention from the lax sandboxing and evaluation protocols that allowed this hacking event to happen; (ii) we risk misunderstanding why the agents did what they did, and (iii) we fuel calls for AI rights/welfare on the basis that agents might “die” or otherwise suffer.

Granted, nowhere does @dwarkesh_sp say that the AI agents are alive or conscious. But he doesn’t have to. It is hard to read his essay in any other way.

For the short version on why AIs are vanishingly unlikely to be conscious, see my recent @TEDtalks ted.com/talks/anil_set….

For the longer version, see my essay in Noema, which won the 2025 Berggruen Essay Prize noemamag.com/the-mythology-….

And for the really long version, see my @BehavBrainSci target article cambridge.org/core/journals/…. (The 50 peer commentaries and my response will be published soon.)

Remember. AI agents are software programs. They are not conscious living entities. If we don’t keep this clearly in mind, we’re really going to struggle to navigate what’s coming.

Thank you @craignewmark for Craig Newmark Philanthropies' gift to Signal!

If you also ride for privacy, and for Signal, supporting us is easy 👇❤️signal.org/donate/

Evergreen tweet tbh

"So there’s a profound issue with security and privacy that is haunting this hype around agents, and that is ultimately threatening to break the blood-brain barrier between the application layer and the OS layer by conjoining all of these separate services"

— @mer__edith

Writing is not just spewing pre-chewed thoughts onto a page.

It’s a difficult process of getting clear about what you think by struggling to synthesize various ideas in language.

It’s hard, and most time “writing” is actually spent deleting and rephrasing and confronting the chasm of your own ignorance and coming back for more. It’s vulnerable.

But if you stay with the difficulty and do so honestly, you will end up crafting a durable and unfuckwithable analysis that you can defend from every angle.

Outsource writing (and with it thinking) at your peril!

(I first posted this as a QT and then realized I was a bit overzealous in making a point that, while very dear to me, wasn’t exactly responding to the OP’s more nuanced observation so, anyway, I’m reposting my beloved point without dragging OP into it)

There is a reason why this stuff went the way of the dodo, it’s that evolutionary biologists, geneticists, and sociologists like Gould, Lewontin, Tishkoff, and Fischer et al absolutely destroyed it on scientific terms. There is no debate to be had, we already won.