Reward engineering defines how AI agents learn. Covers reward functions, dense vs sparse rewards, RLHF reward models, value functions, and reward hacking.
AlphaEvolve and OpenAI Codex: Two New AI Coding Agents Explained
AlphaEvolve is Google DeepMind's 2025 evolutionary coding agent for algorithm discovery. OpenAI Codex is a cloud AI software engineer. Both tools explained.
we discuss Elastic Reasoning, Continuous Thought Machines, ZeroResearch and Absolute Zero plus a fascinating selection of news, papers and things to read
🦸🏻#18: Multi-Agent Systems (MAS): How Collective AI Intelligence Works
Multi-agent AI systems explained: from the 1980s history of MAS research to latest on MARL, LLM-powered agents, swarm robotics, and smart infrastructure.