Extreme Programming with AI?
summary
Concrete claims, measurements and tools dominate: experiments with Claude (including 3.7 and capybara hints) and workspace agents show a rise in copy paste and added code labeled AI slop, GitClear reports copy paste rising from 8% to 12% and added code up 7 percentage points, DORA finds 90% adoption and 80% who feel faster but widespread mistrust, and controlled studies show at-best a 12.5% speedup for skilled users while delivered features become more unstable and prototypes and multi‑agent subagents actually get built.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 3:53AI and extreme programming
- 8:10AI slop and code churn stats
- 11:41DORA report on AI adoption
- 15:56Builder mindset and cognitive offloading
- 18:04Skill erosion with AI agents
- 20:13AI agent limitations
- 23:23AI impact on productivity
- 26:33Extreme Programming values
- 31:47XP practices with AI assistance
- 34:33Multi-agent systems and supervision
- 37:03Prototyping with AI feedback
- 39:10Hype versus reality in AI
- 40:57Q&A
related talks
01h 02m10 Tips To Level Up Your AI-Assisted Coding
26mFrom autocomplete to agents: AI coding assistance state of play
30mBeyond the Hype: Using AI Coding Agents Today
27mSoftware engineering with GenAI
01h 01m10 tips to level up your ai-assisted coding
26mCoding with AI
26mAI Agents as Your Co-Developer
01h 00m10 tips to level up your ai-assisted coding