'ink': an adaptive programming language with proof-backed rules that lets AI optimize itself in a loop
_AustinCalvert_ · x · 2026-10-17
Developer Jungle Silicon unveiled ink, an experimental "adaptive" programming language with an unusual design:
- Compiler machinery lives in a database, where each rule is backed by proofs of equivalence, so AI can optimize code in a self-improving loop;
- Workloads are automatically split across js / wasm / webgpu based on measured performance;
- A key optimization technique is "loosening" what counts as a valid program: exact values can become one of a few candidates, a range, or unordered — developers declare which constraints must hold and which can be unfrozen.
The author says some ideas were adapted from Cheng's proposals. The project links are in the post.
More from coding & agent
- AutoSep cuts agent reward hacking from 69.9% to 12.1% via weak-to-strong verifier hardening — Stanford · 2026-10-17
- Free open-source course teaches harness engineering with 14 lectures and 8 hands-on agent projects — techNmak · 2026-10-17
- Anthropic engineer: Claude Slides is now how I make slide decks — bcherny · 2026-10-17
- Dev built soft-body physics engine with Claude's help, claims it tops BeamNG — grandprixlegends · 2026-10-17
- tldraw ships Codex plugin for pair-programming on a live multiplayer canvas — max__drake · 2026-10-17
- Open Co-Scientist: open-source recreation of Google's AI Co-Scientist with Elo-ranked hypothesis evolution — Haunting-Stretch8069 · 2026-10-17