
NEW EPISODE: @Meta reads your brain with Brain2Qwerty. @AnthropicAI releases Claude Science, and Co-founder and Chairman of @arena Ion Stoica joins us. @istoica05 is the Berkeley professor behind an absurd run of systems that basically run modern AI: @ApacheSpark (the standard for big data), Ray (what OpenAI used to train ChatGPT), and @vllm_project (the most-used LLM inference engine). He also co-founded @databricks and @anyscalecompute - and now chairs Arena, the internet's default scoreboard for AI models. 0:00 The AI memory squeeze 2:55 Brain2Qwerty: Meta Reads Your Mind 9:38 Anthropic Releases Claude Science 15:58 Why Every Lab Wants Pharma 20:52 How Chatbot Arena Was Born 26:58 How Does Arena Actually Rank AI Models? 31:11 Subjective vs Objective AI Quality 33:49 How Arena Evaluates AI Agents 37:54 How Does Arena Stay Neutral? 43:31 Can AI Become Unmeasurable? @Machiz @lightspeedvp
