Figuring out where AI fits (www.schuetzler.net)

A practical, experience-driven take on “where AI fits” argues that students and early-career technologists should avoid outsourcing core learning to GenAI while embracing it as a productivity and creativity multiplier. The author warns that using ChatGPT or other models to do schoolwork robs you of training your own brain — practice, repetition, and problem-solving — but recommends systematically experimenting with GenAI (they point to a 30-day idea list) to build intuition about when it helps and when it fails. They also advise paying for coding-focused tools (ChatGPT/Codex, Claude Code, GitHub Copilot) to use as a pair-programming buddy for personal projects rather than for coursework meant to teach fundamentals. Concrete examples show the practical technical payoff: a Cloudflare-DNS management site built with Elixir/Phoenix, a Python pipeline using Whisper to transcribe class audio and ChatGPT to synthesize summaries, and a local web UI to browse student intro videos. The broader implication is that GenAI lowers implementation barriers — your limiting factor becomes spec’ing and iterating on desired behavior, not raw coding skill — which highlights the growing need for human-in-the-loop design, prompt engineering, structured testing, and ethically mindful use in education and engineering workflows.