#ai-coding
Wiki 21
- acai.sh blog Single-author blog from the developer of acai.sh, on spec-driven AI-assisted development
- Acceptance Criteria IDs (ACIDs) Stable numeric tags on requirements that agents reference back from code and tests, turning the spec into a navigable index of where requirements live
- Addy Osmani blog Engineering manager at Google Chrome's developer-tools team; long-form essays on web performance, software craft, and the cognitive economics of AI-assisted coding
- An Empirical Study on Strong-Weak Model Collaboration for Repo-level Code Generation CMU compares strong-weak model splits on SWE-bench Lite; the strong model writing first wins, and no strategy tested is a review stage
- Codex-maxxing jxnl's operating loop for Codex — durable threads, voice input, steering, file-backed memory, Heartbeats, Goals, and the side panel as where work lives
- Cognitive Debt MIT's framing for the trade where LLM-assisted work saves effort today and pays for it in critical-thinking capacity tomorrow
- Constraint Decay LLM coding agents look strong on loosely specified tasks but degrade sharply as production structural constraints accumulate
- Constraint Decay: LLM Agents in Backend Code Generation A benchmark showing coding agents lose ~30 assertion-pass points as production structural constraints accumulate
- Defining AI Psychosis, Part 2: Prolific AI Psychosis Psychiatrist Jeff Clark defines prolific AI psychosis as heavy AI output with little real value and a lost ability to judge one's own work
- Don't Outsource Learning Addy Osmani's argument that the default AI coding loop quietly trades comprehension for completion, with three converging studies and a posture-shift prescription
- I'm Going Back to Coding by Hand (Ask HN) An HN user abandons six months of Claude-built features on their own app after realizing they no longer knew how the new code worked
- Improving Code Generation via Small Language Model-as-a-judge Fine-tuned sub-5B judges pick correct Java code about as well as GPT-4.1-mini; small generator plus judge rivals its 8-33B sibling in 4 of 5 families
- Know Thine Enemy (Ko) Amy Ko's 3-month experiment using Claude Code as an LLM skeptic — it worked, and cost her focus, joy, money, and values
- Local AI is not Opus Alex Ellis on why "local Qwen is near-Opus" is wrong, but a tuned local model still earns its keep in a small software business
- Measuring AI Coding Productivity The recurring methods failures that make most claims about whether AI coding tools help unreliable — proxy metrics, missing controls, novelty, selection, systems confusion
- My AI-Assisted Workflow Barbero's seven-step workflow where planning happens before code and AI stress-tests thinking rather than replacing it
- Programming Is Art (Orchid) Orchid splits developers into builders who code for money and true programmers who code for its own sake and will never hand it to AI
- Software Engineering Is About Managing Complexity (hack8s) hack8s.com on why cheap AI-generated code makes engineering judgment and real ownership of a codebase scarcer instead of less needed
- Specsmaxxing — Acceptance Criteria as the Primary Artifact acai.sh's author on numbered acceptance-criteria IDs as the missing link between agent-written code and a stable spec
- SWR-Bench: LLM Code Review Comment Generation on Real Pull Requests 1,000 real PRs with full repo context; the best LLM review setups reach about 20% F1, dragged down by false positives, and pooling several reviews helps
- Twelve Ways to Be Wrong About AI-Assisted Coding Greg Wilson's catalog of 12 measurement errors in studies of whether AI coding tools work, each mapped to a known research-methods failure
talks 2
- PL Education in the Age of AI Shriram Krishnamurthi on teaching PL to the 90% who won't do research, confidence over proof, and what agentic coding leaves CS to teach
- Software Fundamentals Matter More Than Ever Matt Pocock argues AI coding rewards old fundamentals — shared design concept, ubiquitous language, TDD, deep modules — and bad code costs more than ever