Observatoire des droits humains (anglais seulement)

AI has a stupid secret: we’re still not sure how to test for human levels of intelligence

(Version anglaise seulement)

par Andrew Rogoyski, Innovation Director - Surrey Institute of People-Centred AI, University of Surrey

Two of San Francisco’s leading players in artificial intelligence have challenged the public to come up with questions capable of testing the capabilities of large language models (LLMs) like Google Gemini and OpenAI’s o1. Scale AI, which specialises in preparing the vast tracts of data on which the LLMs are trained, teamed up with the Center for AI Safety (CAIS) to launch the initiative, Humanity’s Last Exam.

Featuring prizes of US$5,000 (£3,800) for those who come up with the top 50 questions selected for the…

Lire l'article complet