Leadership in AI-assisted engineering
summary
Claims are concrete: measured averages show a 2.6% change confidence gain and about a 1% decrease in change failure rate, while per-company results swing ±20%. Leadership tactics that move metrics include clear AI policies, mandatory training, and a measurement framework using PR throughput, cycle time, and change failure rate. Real rollouts include Morgan Stanley’s DevGen AI saving 280,000 hours annually, Fair’s agent doing 3,000 reviews a week, Canva generating PRDs, and Spotify automating ~90% of internal incident handling.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 0:48GenAI impact on productivity
- 2:52AI adoption challenges
- 3:34Key factors for AI success
- 4:17Fear of AI as a tool
- 5:21Psychological safety in AI teams
- 6:25Measuring AI impact
- 8:10Developer experience framework
- 9:57AI adoption versus outcomes
- 10:17Compliance and trust in AI code
- 11:39Human review and bias
- 12:21Education for AI skills
- 13:25Barriers to AI adoption
- 14:31AI beyond code generation
- 15:34SDLC success stories
- 16:18Next steps and resources
related talks
29mLeadership in AI-assisted engineering
22mThe AI Productivity Paradox in Engineering
27mThe state of AI in software development: Insights across 400+ organizations
45m"How AI improved our performance by 20%" and other bedtime stories
01h 01mAre Your Engineering Teams Prepared for the GenAI Revolution
32mAI and Engineering Closing the Gap Between Expectations and Reality 1
22mHow Can Engineering Leaders Measure the Impact of GenAI
44mSpeed Is Easy, Trust Is Hard: AI's Real Impact on Developer Productivity