"11 months from level 1 to level 1,000": dev reflects on agentic coding's wild pace
Dimillian · x · 2026-10-09
A developer recalls watching a @Dimillian agentic coding demo a year ago and feeling insecure using the then-weak coding agents. He notes the progress from "level 1" agents to today's capabilities happened in just 11 months — a snapshot of how fast AI coding tools have iterated.
More from coding & agent
- REMORY adds soft residual memory tokens to context compaction, hitting near full-context scores at 5.2% of input — Hanchen Xia · 2026-10-09
- Hermes Bot Mode demo: auto-generate a team of specialist agents from project requirements — Teknium · 2026-10-09
- Using Claude Code on NASA TESS data, researcher finds two Earth-sized exoplanet candidates — paraschopra · 2026-10-09
- Skillry launches Deck Skills: agents now generate designer-quality, editable decks — yihui_indie · 2026-10-09
- Dev proposes hybrid harness routing cheap classifier models with LLM fallback to cut AI bills — Efficient-Path4172 · 2026-10-09
- Non-technical recruiter builds Claude Code agent that sourced 40 candidates in 6 weeks — Jags2ooo · 2026-10-09