How Agents Quietly Break Architecture
German dude meandering about how AI agents can quietly break software architecture, why passing tests is not always enough, and how formal models in Lean can help keep the intended meaning of a system close to the implementation. 00:00 Introduction 02:34 Example Project 03:59 High-Level Architecture 05:21 Why "Green Tests" Can Lie 10:26 LEAN Intro 13:10 What The Agent Did 13:55 Rust Implementation 17:47 Spec Driven Development #softwaredevelopment ai #agenticcoding #formalmethods #leanlang #programming #codequality #aiengineering
▶︎
Why AI Has Failed to Take Your Job Since 1976
▶︎
Recursive Self-Improvement
▶︎
What Happened to AI Assistants? - The END of OpenClaw
▶︎
The Craziest Coding Contest Ever
▶︎
Reinventing Entropy | Compression is Intelligence Part 1
▶︎
Google Did The Impossible
▶︎
The First Domino Has Fallen...
▶︎
If Prime Numbers Become Increasingly Rare, Then Why Do They Keep Showing Up In Pairs?
▶︎
The Only Unbreakable Law
▶︎
How Is a Memory Stored Inside Your Squishy Brain?
▶︎
I Think We're Losing Control Of AI
▶︎
Karma Just Hit Adobe. Hard.
▶︎
Agent Skills or MCP in the era of Claude Code?
▶︎
I Think They Are Lying To You
▶︎
China's 1.4nm Breakthrough Terrifies America and Taiwan
▶︎
"Software Fundamentals Matter More Than Ever" — Matt Pocock
▶︎
Teenager Disproves 4 Decades Old Belief in Computing
▶︎
The Hidden Cost of AI Coding That's Destroying Engineering Teams
▶︎
Harvard just discovered what AI actually is
▶︎