Reviewing this amazing Lex Fridman podcast where he interviews the CEO, Dario Amodei and folks from Anthropic. They start with the scaling problem where one has to linearly dial up data, compute and training to improve the intelligence of the network. Lex asks very interesting questions (What is the limit of compute? What are the … Continue reading Mantra M5 Thesis Brief 64 – Scaling AI using LLM, LWM and LGM