AI agents 101
summary
Claims that agents are not just chat models but systems made of LLM + memory + tools that act autonomously, decide which tool to call, and execute code and web searches to complete tasks. Demonstrates booking a flight (Agadir→Paris, dates given) where the agent chooses a flight tool, runs code to query APIs, aggregates options, and returns a final answer. Notes common frameworks that generate and run code, warns about drift from predefined flows, safety risks, and practical tradeoffs like cost and reliability.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:18What is an AI agent
- 4:45How AI agents work
- 8:45AI agents vs models
- 13:03Memory in AI agents
- 16:34AI agent frameworks
- 19:10Philosophy behind AI agents
- 23:19Future of AI agents
- 24:46Resources and conclusion
related talks
40mAI Agents in Practice
26mAI Agents: How Do They Work and How to Build Them
49mAI Agents: Collaboration in Real-Life Applications
40mAI Coding Agents and how to code them
34mAI Agents: Transforming Autonomous Decision-Making
13mBuilding AI Agents is easy
01mAI Agents are the Future of Coding?
40mAgentic AI: The Next Evolution of Automation