Practice · 2023–now · AI Present
Shadow AI / paste-into-ChatGPT
Prod data into consumer chat because the approved tool is slow. GDPR with better autocomplete.
Shadow AI failed as policy-by-hope because convenience beats blocked portals. It sticks as a governance problem wherever sanctioned tooling lags real workflows. The costly fad was pretending training slides stop paste; the durable fix is usable approved paths and data classification that engineers can follow under deadline.
Cost of the fad
Employees routed tickets, logs, and customer PII through consumer chat before InfoSec finished procurement. Compliance scramble followed the demos.
Case studies
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
Practice · 2021–now
AI pair programming
Autocomplete that understands the file — now table stakes. Generation outran review capacity; ownership is the bottleneck.
Practice · 2018–2025
Low-code will replace eng
Drag-and-drop apps for the enterprise backlog. Useful for internal tools; fantasy as a full replacement for engineering — a pitch later absorbed by AI demos.
Practice · 2023–now
Prompt injection as afterthought
Ship the chatbot, bolt on security later — OWASP's new category, same old afterthought.
$ Customer-facing chatbots shipped with tool access and "security in phase two." Indirect injection via docs and tickets became the new XSS.
Practice · 2023–now
Local inference
Run the model on your laptop or rack. Privacy, cost control, and the refusal to send every token to a vendor.