Braid

Coding is solved, the rest isn't

Listen Later

Boris Cherny says coding is solved for the coding he does — and almost everything else in today's research is a study of the parts that aren't. A new coding leaderboard with an accusation, the end of the "software engineer" title, the craft of delegating to an agent, and three papers on the ways agents quietly break: introspection, aging, and memory. Plus running a trillion-parameter model in your house, the labs' jobs split, and a developer who's tired of talking to AI.

  • DeepSWE crowns GPT-5.5, and accuses Opus of cheating — what looks like a loophole may just be a model recovering the answer from git history.
  • The end of the software engineer, in the first person — Cherny in Platformer and Steven Levy in Wired on the agent boom and its hazards.
  • What the best agents share, and how to drive one — Flinn AI's four patterns alongside a practical Claude Code daily-driver guide.
  • Can the model actually tell when it's unsure? — a reality check on LLM introspection and self-reported confidence.
  • Your agents are aging — AgingBench, MemFail, and rethinking agent memory as a state trajectory.
  • Running the frontier in your own house — EXO Labs on local inference economics and the 100x still left.
  • The labs can't agree on the jobs — Anthropic vs OpenAI, with Hassabis calling 2026 a practice run.
  • I'm tired of talking to AI — a developer on people forwarding AI answers they never read.
    • ...more
    View all episodesView all episodes
    Download on the App Store
    BraidBy Lenar Kess · Damra Vol