Weekly critical commentary on artificial intelligence, from the philosophy of technology and with a focus on the Global South.
UN General Assembly week: twenty-two leaders call for an institution able to act on the frontier and the three countries that host the labs do not sign, the companies ask the …
The first encyclical on artificial intelligence does with data and algorithms what Rerum Novarum did with the factory: it does not ask how to use them well but whom they belong to. …
MIT documents "cognitive surrender" and answers with oral exams, portfolios, mentoring and more undergraduate research. Every one of those recommendations is paid for in teaching …
The week the frontier announced it would slow down and the market did the arithmetic, while UNESCO presented a readiness questionnaire in Riyadh: two ways of governing the same …
The week the risk of extinction jumped from online forums to front pages, to the US Senate and to an interview in Buenos Aires: a threat that by definition includes all of us, …
Commentary on Terence Tao’s video about the paradox of accelerating science with AI: why his final point —the defence of curiosity-driven research and of the community that …
The week OpenAI released the model it had halted for being dangerous, New York banned what it cannot see, and the detector used to measure how much of the web is AI-written turned …
Claude formalized the first computer-checked proof of Fermat's Last Theorem in eleven days. The theorem is not the interesting part: that domain had a cheap, independent, hostile …
Commentary on Maya Makarovsky’s TEDxMIT talk on human connection as the decisive skill against AI: what MIT’s own format says that the talk does not, and what question it leaves …
The week protecting minors meant identifying every adult, a philosophy journal banned what it cannot detect, and Amazon finished scanning the books it then throws away.