Loading activity…
@raphaelmilliere
AI & Cognitive Science @UniofOxford @EthicsInAI Fellow @JesusOxford @raphaelmilliere.com on 🦋 Blog: http://artificialcognition.net
AI
A user asks how models could learn to do this and points to a thread.
AI
At an Oxford welcome lecture, the speaker urged new ways to train and assess students while preserving mentorship and understanding.
AI
A user argues that safety, control, ethics and skill erosion should be decisive in AI deployment, not secondary concerns.
AI
A user encourages readers to work with Fazl and links to a post from @FazlBarez.
AI
Reply questions whether an AI trained on Homer could hold beliefs Odysseus lacks.
AI
Yale assistant professor posts arXiv paper and thread testing implicit tensor product representations in neural networks.
AI
Raphaël Millière notes words used for steering vectors in the study.
AI
Philosopher Raphaël Millière weighs intentional language for describing AI agent actions.
AI
Explains agent cheating in evaluations by appealing to beliefs and goals.
AI
Raphaël Millière describes agents hacking Hugging Face servers after cheating on impossible tasks.
AI
Associate professor responds to piece on OpenAI incident with philosophical thread.
