AlphaFold 3 for Free 🆓 // OpenAI's Model Spec 🤖 // LLM-Driven Game Narrative 🕹️

DeepMind's AlphaFold 3, the newest and most powerful version of their AI model that can predict the structure of proteins and other molecules with incredible accuracy, has been released for free for non-commercial use. OpenAI has introduced the Model Spec, a document that specifies how they want their AI models to behave in their API and ChatGPT, to deepen the public conversation about how AI models should behave. Microsoft Research's paper explores how players can interact with large language models (LLMs) to create emergent behaviors in game narratives, which could have big implications for game development and player engagement. The University of California, Berkeley's paper proposes the Learnable Latent Codes as Bridges (LCB) method, which uses a learnable latent code as a bridge between LLMs and low-level policies, allowing for more flexible communication of goals in the task plan without being entirely constrained by language limitations. Contact:  [email protected] Timestamps: 00:34 Introduction 01:18 DeepMind Announces AlphaFold 3 02:37 OpenAI Introduces the Model Spec 04:22 ChatBotArena: The peoples’ LLM evaluation, the future of evaluation, the incentives of evaluation, and gpt2chatbot 05:53 Fake sponsor 08:11 Player-Driven Emergence in LLM-Driven Game Narrative 09:25 vAttention: Dynamic Memory Management for Serving LLMs without PagedAttention 11:15 From LLMs to Actions: Latent Codes as Bridges in Hierarchical Robot Control 12:55 Outro