I decided to go to DC next week to discuss directly with policymakers. Not sure about the impact it will have, but with everything going on, it seems like a good time to speak more about open-source AI, transparency, the concentration of
POLICY
-
Lawmakers warn airlines of AI pricing targeting personal pain points
By
–
1/ Airlines are moving fares to AI that prices by demand, your device, and your location in real time. One carrier planned to set 20% of fares this way, in what lawmakers called pricing aimed at your personal "pain point." It denies using personal data. The rest are following.
-
Interview with Roland Lescure on France’s place in AI
By
–
The full interview on France's place in the global AI race with the Minister of Economy @RolandLescure is available on my YouTube channel! https://youtu.be/BF4y9O89gJ4 Here are all the topics covered: CHAPTERS
00:00 Introduction
00:21 €655M for AI -
Global AI rules: labs, governments, and who is missing?
By
–
My question: Should a handful of labs and governments write the global rules for AI? And who's missing from that table?
-
Anthropic asks US to lead AI after government pulls models
By
–
The honest tension, Anthropic asked the US to lead global AI 5 days after that same government pulled its top models offline. No binding deals, but the labs want a seat at the table where the rules get written.
-
Altman proposes international AI testing standards forum with global leaders
By
–
Altman pitched an international AI-testing-standards forum. Canada's Carney said yes, the US could lead. Mistral, Cohere, Meta, Salesforce + labs from India, Japan, Italy all at the table. France framed it as global, not American.
-
Amodei’s allied frontier AI access and joint defense proposal
By
–
Amodei's proposal – structured access to frontier models across allies, chip/component trade that cuts out China, joint defense on cyber/bioterror/intelligence risk. Allies setting terms together instead of racing each other off a cliff.
-
AI rivals agree: US should lead global AI alliance at G7
By
–
At the G7 this week, the CEOs of Anthropic and Google DeepMind sat at a closed-door lunch with Trump and a dozen world leaders and pushed the same idea: let the US lead a global AI alliance. Two labs that compete on everything. Agreeing on this. The rules for AI are being written
-
Who designs RL agent’s training environment: practitioner or policy?
By
–
Who should design the training environment for an RL agent, the practitioner or the policy itself? RL pipelines for LLMs usually rely on manually redesigned environments between stages, with practitioners guessing which configuration will best improve the current policy. This
-
AI as atomic weapons, unpatchable jailbreaks lead to this outcome
By
–
To be fair, if one was to repeatedly describe AI as equivalent to atomic weapons, then saying that patching all jailbreaks is impossible, I totally get how we'd end up here