AI and APIs in 2026
summary
Gartner claims software shifts to tiny teams of two combining one technical lead and one domain expert, using LLM tools like Claude, OpenAI Codex, Cursor and Cloud Code; current development is command-line agents with humans in the loop and a predicted phase where agents orchestrate themselves. Interest in AI gateways rises from single‑figure inquiries in 2023 to much higher in 2025; vendor work includes model routing, cost/token controls and sandboxes with demo keys. New category is agent management platforms with dashboards, marketplaces and AMP-driven governance.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:30AI’s impact on API trends
- 4:43Two phases of AI-driven dev
- 7:14Measuring AI’s productivity
- 9:23AI gateways in API management
- 12:54Agent management platforms
- 15:46Risks of API bypass by AI
- 18:13Designing APIs for AI agents
- 22:26Summary
related talks
28mAI and the state of APIs in 2026
26mAI & the state of APIs in 2025
10mInvesting In AI Era
47mAI and the Future of Software Development
39mAI as a Catalyst, Not a Replacement
01h 06mAI in the Trenches: Engineering Real-World Systems
27mWill AI (Finally) Replace Developers?
28mTransforming Internal Developer Platforms with AI: Accelerating Innovation