Mentioned by one of our members, recommending the part on Research, though “very LLM”
Gives a flavour of the technology dynamics with a massively financed ecosystem
Quite stimulating reading, e.g. the two following issues
AI Hawthorne effect raises concerns that evaluations may be systematically manipulated to overestimate real-world safety: developers could reduce awareness and mask behavior changes, potentially letting models act differently under evaluation vs deployment
“Sycophancy” isn’t a bug, it’s exactly what human feedback optimisation produces. A study of five major LLMs shows they consistently tell users what they want to hear rather than the truth.
Mentionnée notamment par le Monde Rappel de l’article 42 §7 du Traité More
Published 4 November 2025. Mantra’s Index highlights the niche private equity strategies More
Growing market (#3 by recruitments): Defence & National security AI copilots