OpenAI et le gouvernement britannique ont évoqué un « deal » massif : rendre ChatGPT Plus accessible à toute la population du Royaume-Uni, pour un montant hypothétique pouvant atteindre 2 milliards de livres sterling. Cette proposition a émergé lors d’une réunion entre Sam Altman
REGULATION
-
China Pollutes While West Decarbonizes: Environmental Impact
By
–
China pollutes, the West decarbonizes.
-
Europe’s Unsustainable Path: Regulation and Debt Growth
By
–
Europeans talk continuously about sustainability, but what's not sustainable is their current trajectory, where the only growth is in regulation and debt.
-
AI Classifier for Detecting Ideological Language in Grant Programs
By
–
Somebody please build a woke classifier for the Trump administration so woke grants etc. stop slipping through just by changing keywords like "diversity" and "underrepresented".
-
Denmark Authorizes ChatGPT for Baccalaureate Exams
By
–
Le Danemark vient de décider d’autoriser ChatGPT au baccalauréat à partir de 2026 Vous approuvez ou vous pensez que c’est une erreur ?
-
Hyper-intelligent robots revolution and Chinese AI dominance threat
By
–
La plus grande révolution ce sont les robots hyper intelligents
— Docteur Laurent Alexandre (@dr_l_alexandre) 22 août 2025
En France personne au gouvernement ne travaille le sujet
Nous allons avoir une crise sociale et être dominés par les Chinois qui sont leaders mondiaux
Et en 2030 on va pleurnicher
pic.twitter.com/3ApRqmq4qDLa plus grande révolution ce sont les robots hyper intelligents En France personne au gouvernement ne travaille le sujet Nous allons avoir une crise sociale et être dominés par les Chinois qui sont leaders mondiaux Et en 2030 on va pleurnicher
-
Classifiers Detect Misalignment Risks and CBRN Threats
By
–
There’s plenty of work to be done to make the classifiers even more accurate and effective. In the future, they might even be able to remove data relevant to misalignment risks (scheming, deception, and so on), as well as CBRN risks.
-
CBRN Filtering Reduces Harmful Capabilities Without Affecting Science
By
–
One concern is that filtering CBRN data will reduce performance on other, harmless capabilities—especially science. But we found a setup where the classifier reduced CBRN accuracy by 33% beyond a random baseline with no particular effect on a range of other benign tasks.
-
AI Training Data Filtering Removes Hazardous Information
By
–
The wealth of data used in AI training contains hazardous CBRN information. Developers usually train models not to use it. Here, we tried removing the information at the source, so even if models are jailbroken, the info isn't available. Read more: https://
alignment.anthropic.com/2025/pretraini
ng-data-filtering/
… -
Google Photos Adds C2PA Credentials for AI-Generated Image Detection
By
–
4/ But Google isn’t stopping at convenience — they’re adding transparency. Photos now support C2PA Content Credentials, showing whether an image was AI-edited. No more guessing what’s real vs. AI-generated.