Documentation and AI: How to write right now
summary
AI excels at crawling and extracting and indexing 80GB corpora, templating predictable outputs like security release notes and SDK translations, and doing polishing work such as refine grammar and linting. Practical examples include enterprise search and knowledge tools—glean alolia unblocked, swimmli, docuatic, GitHub, Clickhelp, Doxygen, Docosaurus, Confluence—but useful systems must export text, metatext, and context and support audit trails. AI fails on novel cross‑context analysis, business‑decision rationale, and human feelings; it helps where tasks are repetitive and predictable.
AI-written summary · about
chapters (21)
- 0:00Intro
- 0:22Heidis story
- 2:44Summary
- 3:21Frustrations
- 3:54Boring
- 4:28Automate
- 5:01Documentation as content
- 6:15Documentation is intensely contextual
- 7:31Contextsensitive help
- 8:33Docs for Developers
- 9:05Robots
- 9:45Ideal AI provider
- 11:38Imperfect world
- 12:51Extract
- 14:00Refine
- 14:51Presentation
- 17:12Essential tools
- 18:29Where does AI help
- 19:57AI cannot understand
- 21:37Give AI documentation a chance
- 23:19Progressive Delivery
related talks
39mWhat I Learned Building an AI Documentation Auditor (And What It Found in 500+ Pages of Docs)
15mAI's Impact on Development, Post-Production & Support Journeys
39mDocs AI Tooling is Better (and Better for Us) than You Think
23mDocumentation in the Age of AI: Why Docs Are a Product
23mAI-native CMS vs vibe coding
49mNext-Gen Documentation with AI
16mInteract with Your Technical Documentation via Generative AI
35mHarnessing AI for API Business Documentation in a Devportal CMS [...]