New Methods of Working with AI.
summary
Argues GenAI shifts SDLC from individual copilot help to agent-driven workflows, with agents built from prompt, planner, executor, memory and deployed as either sequential or autonomous orchestrations. Reports concrete results: a unit-test generator hits ~70% coverage from zero on test repos, and an AI coding pipeline scores SWEbench 39.6% (leader 53%) using Claude 3.5; shows a working demo that migrates a Spring Boot app to AWS serverless with generated Lambda code, CI/CD and a PR — built in 2–3 days — while warning about instability from model changes and data‑leakage risks.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 3:18Speaker and AI focus
- 5:04Generative AI as game changer
- 7:17Human-machine challenges
- 13:42Five waves of AI impact
- 18:03Productivity gains from AI
- 24:32AI agents and components
- 29:35AI workflows sequential and autonomous
- 35:48AI workflows practice and benchmarks
- 40:08Live demo AI migration
- 45:48Adoption challenges and trends
- 49:28Q&A
related talks
26mAI Agents as Your Co-Developer
27mSoftware engineering with GenAI
21mGen AI disrupts SDLC. What does it mean for developers? The AWS approach | AWS Developer Day 2025
59mFrom Pair to Peer Programmer
30mHow AI Agents Will Change the Future of Programming
21mRevolutionizing software creation: Harnessing the power of GenAI
25mFrom AI Chaos to Production Ready: Using AI Agents to Improve Your Code Quality
41mGoing beyond copilot with AI Agents