Programming spent forty years climbing away from the machine β garbage collection, ORMs, dynamic typing, magical frameworks β trading runtime cost for human comfort while a person was at the keyboard. If LLMs are writing most of the code, the next generation of languages won't optimise for what's pleasant to type, but for what an agent can synthesise, inspect, test, and repair without anyone stepping in.
The full post argues this isn't "we all go back to C" β it's that visibility at the call site, not abstraction level, becomes the new design axis. SQL keeps its abstraction; ORMs that materialise seventeen joins from one expression don't. Walks through the AI adoption tax on new languages, why Python complicates the picture, and ends on a falsifiable bet about which TypeScript and Java backends gain share by 2029.
Originally published at andreasbergstrom.dev β read the full post there.
United States
NORTH AMERICA
Related News
What Does "Building in Public" Actually Mean in 2026?
19h ago
The Agentic Headless Backend: What Vibe Coders Still Need After the UI Is Done
19h ago
Why Iβm Still Learning to Code Even With AI
21h ago
I gave Claude a persistent memory for $0/month using Cloudflare
1d ago
NYT: 'Meta's Embrace of AI Is Making Its Employees Miserable'
1d ago