Generative Value Streams: When Code is not a Constraint
summary
Claim: generative AI removes coding as the delivery constraint and delivers productivity far larger than open source. Evidence: code agents like GitHub Copilot are already in use; three agent tiers — code, value-stream, strategic — automate refactoring, roadmap and org design. Data from 3600 value streams across 34 organizations shows a 10x mismatch between demand and capacity, 40% wasted effort, 30–40% of flow time stuck post-deploy, 80% neglect tech debt and 90% measure outputs not outcomes. Actions: adopt copilots, measure flow, shift from project to product, form an AI committee.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:44Generative AI productivity
- 4:15Next tech revolution
- 6:26AI productivity multipliers
- 8:32Code agents and constraints
- 10:16Value stream agents
- 12:24Strategic agents
- 13:49Organizations not ready
- 15:52Strategy, value streams, code
- 16:56Call to action
- 18:20Measuring impact
related talks
34mBeyond CodeGen: How AI & Machine Learning Are Transforming DevOps
21mRevolutionizing software creation: Harnessing the power of GenAI
33mAI assisted software delivery
38mFrom code to culture: How AI is reshaping the engineering organization
30mThe AI Revolution in Coding
20mOutput to Outcome: An Operating Model for the Age of AI
38mGenerative DevOps
53mThe Human Side of AI: Unlocking Opportunity in the New Bottlenecks of Software Development