Barr Yaron says AI engineering in 2026 is defined by agents, cost, and evals
AI Engineer · youtube · 2026-07-22
Barr Yaron’s 2026 State of AI Engineering talk says AI engineering has moved from experiments to operating constraints that shape the whole stack.
- The median new engineer now has nearly as much AI experience as a 10-year veteran, which is changing team composition.
- Cost has become a first-class engineering constraint, not an afterthought.
- Agents are no longer just drafting or summarizing — they are taking actions inside systems.
- Evals remain the recurring bottleneck in the stack, year after year.
- The talk also highlights blurrier role boundaries and a build-vs-buy trade-off that is increasingly central to AI teams.
More from coding & agent
- marimo Glance: Turn GitHub Python Code into Live Interactive Notebooks — S_Conradi · 2026-07-22
- Cognition says Devin Outposts can run on a Mac mini, VM, GPU box, or Kubernetes — GavinSBaker · 2026-07-22
- Building a Knowledge Graph Without a Graph DB: 1000x Cheaper Than GraphRAG — TheRedfather · 2026-07-22
- FactoryAI gave back its first millions, then shipped Droid CLI two years later — matanSF · 2026-07-22
- Hermes Agent Refactoring Proposal: Decoupling via Event Bus and Monorepo Slicing — Promptmethus · 2026-07-22
- Show HN: a 6.2 MB pure-Go terminal command palette with no fzf dependency — MarinhoD · 2026-07-22