Devin 2.0 and the Future of SWE

on YouTube ↗16m

" summary

Claims that code agent capability doubles about every 70 days, enabling 16–64x yearly gains, and shows concrete progress from line completion to full project work. Built features include playbooks for stepwise migrations, a memory/knowledge layer, repo snapshots with remote VMs to run CI and linters, deep wiki and search for codebase exploration, IDE and Slack integration, and asynchronous self-testing; by June the system can autonomously clear large backlogs and deliver multi-file PRs.

AI-written summary · about

" chapters (13)

1. 0:00Introduction
2. 0:43Moores Law
3. 2:09SWE Tiers
4. 3:31Repetitive migrations
5. 5:12playbooks
6. 6:27knowledge or memory
7. 7:09a level of task
8. 8:13bug requests
9. 8:44working with context
10. 10:39Improving the architecture
11. 11:34Improved codebase intelligence
12. 13:41Confidence
13. 14:33Conclusion

" related talks

• 06h 22mUsing AI to Enhance Developer ProductivityTyler Leonhardt, Martin Woodward, Olivia Guzzardo, Kai Maetzel, Brigit Murtagh, Connor Peet, Logan Ramos · AI Dev Days 2025
• 16mCoding Reinvented How AI Agents Are Shaping the Future of Software DevelopmentArchie Sharma · Devworld Conference 2025
• 29mDevTools for the decade to comeKenneth Auchenberg · revo.js 2019
• 14mAgent-Driven Software DevelopmentEno Reyes · SuperAI Singapore 2025
• 18mCoding Reinvented: How AI Agents Are Shaping the Future of Software DevelopmentArchie Sharma · Great International Developer Summit 2025
• 26mAI Agents as Your Co-DeveloperAmer Sheeny · AI Agents Conference 2026
• 35mGoing Beyond Copilot with AI AgentsAlex Shershebnev · PyData Amsterdam 2024
• 36mThe evolution of Software engineeringTheophane Gregoir · Devoxx France 2026