Practice · 2021–now · AI Present
AI pair programming
Autocomplete that understands the file — now table stakes. Generation outran review capacity; ownership is the bottleneck.
Copilot-style tools stuck because typing boilerplate is not where senior engineers create value. By 2026 the tool is assumed; the unpaid bill is review, security, and knowing which prompt wrote the auth middleware. The fad layer is assuming correctness; the tool layer is real.
Case studies
- AI codebase rewrite pilots that shipped nothing — Six-figure LLM API and SI costs; production cutover: zero
- Low-code to replace engineering — then hire engineers — Seven-figure platform licenses; shadow IT integration debt
- Vibe-coded auth middleware nobody could explain — Sev-1 incident; emergency rewrite; weeks of audit and customer trust repair
- Shadow AI — tickets and logs in consumer chat — Compliance scramble; vendor DPA retrofit; partial customer notification
Patterns
Context
Demos consolidate; evals remain
AI pair programming changed how code is typed faster than how it is reviewed. Agent frameworks and standalone vector stores sorted into demos versus durable plumbing; mid-market RAG folded back into Postgres. The permanent layer is familiar: evals that gate deploys, model routing for cost, tool protocols instead of plugin snowflakes, and humans who own production. Autopilot rewrites and vibe-shipped auth middleware are still big-bang migrations with better slides — and a longer on-call.
Compare with
Related
Framework · 2023–now
LLM app frameworks
LangChain-class glue gave way to vendor SDKs and thin wrappers. The durable pieces are boring: evals, retrieval, and product UX.
Language · 2012–now
TypeScript
JS with a typechecker that conquered the ecosystem. The rare fad that became the floor.
Practice · 2023–now
AI will rewrite the codebase
Big-bang rewrite, now with a chatbot — then with agents. Same failure mode; prettier slides.
$ Exec decks promised autonomous migrations "by next quarter," then rebranded as "agentic modernization." Teams that skipped tests, ownership, and incremental strangler patterns bought expensive demos and fragile diffs.
Practice · 2024–now
AI review bottleneck
Copilot ships PRs faster than teams can review. Generation outran ownership and security review — the real 2026 invoice.
Practice · 2024–now
Vibe coding
Accept AI output on feel, skip the diff, hope tests exist. Fast demos; CVEs and nobody owns the middleware.
$ Ship-to-prod without reading the diff became a flex until incidents revealed nobody knew which prompt wrote the auth middleware. Review debt arrived in one weekend demo.