Every's head of evals shares 13 beliefs on writing with AI after finishing second O'Reilly book
every · x · 2026-09-24
Mike Taylor, head of evals at Every and co-author of "Prompt Engineering for Generative AI" (O'Reilly), lays out the 13 beliefs behind his AI writing practice—how he went from calling outputs slop to publishing under his own name.
Key points: he just finished his second O'Reilly book (which bans AI writing) and says it's the last he'll write manually—traditional writing feels like using a typewriter when you have a word processor. He's experimented since GPT-3 in 2020 and by GPT-4 had a process that passed the Turing test with friends. Core principles include putting more of his own material into the prompt than he asks the model to produce, and keeping social posts entirely his own. His test: could he defend the ideas if a reader showed up to argue?
Related event: Every's Review Lead Shares Full AI Writing Workflow(2 posts)→
More from Apps
- Lovable Chat tutorial: voice, file creation, and auto-posting to X via Metricool MCP — MyCreativeOwls · 2026-09-24
- Lovable Chat hands-on: voice, file creation, and auto-posting to X via Metricool MCP — MyCreativeOwls · 2026-09-24
- Solo dev launches Pixora, an AI app turning pet photos into custom stickers — jakubzeg · 2026-09-24
- DeepSeek Harness Quietly Ships Official GUI Client for Windows and Mac — vista8 · 2026-09-24
- Quantum Odyssey: a 9-year passion project teaching quantum computing via visual puzzles — QuantumOdysseyGame · 2026-09-24
- MDFlux: open-source local-first tool turns scanned PDFs and Office files into AI-ready Markdown, 6x fewer tokens — tom_doerr · 2026-09-24