An agent writes code quickly. It builds what you said, and it misses what you meant.

I drew four short chapters from the Manning mini-ebook Spec-Driven AI Development: From Intent to Trusted Code. The narration is in English, with subtitles. The ideas belong to Hari Krishnan, Leigh Griffin, and Tomasz Lelek and Artur Skowroński.

Not watched yet. Saved to watch later.

Video from Den Delimarsky.

Not read yet. Saved to read later.

Post by Anthropic.

Not watched yet. Saved to watch later.

Video from Georg Šumailov.

Not watched yet. Saved to watch later.

Video from Anton Arhipov.

Not watched yet. Saved to watch later.

Video from Frank Coyle.

Not read yet. Saved to read later.

Post by Justin Poehnelt.

Not read yet. Saved to read later.

From the Thoughtworks Technology Radar.

Not watched yet. Saved to watch later.

Video from Samin Yasar.

Not watched yet. Saved to watch later.

Video from Jadi Mirmirani.

Video from Maryam Sadeghi.

Takeaways

• AI agent elements: LLM, tools, goal, loop, memory.
• Types of memory
  • Working memory: the current conversation, like RAM
  • Procedural memory: how it should behave. Rules and instructions (skills)
  • Semantic memory: stable facts we want the agent to always know
  • Episodic memory: a timestamped history of past events and conversations

These memories are stored in a database.

The cost gets high. Distill: summarize and pull out a few important things, then store them in the database.

Welcome to my first blog post! In this post, I’ve shared an AI-generated podcast created by NotebookLM exploring the book “The Professional Scrum Team”.

If you’re curious to hear how NotebookLM discusses team dynamics and Scrum principles, you can listen to the podcast below: