Skip to content
Menu
WorkflowSkills, tmux and the daily loopToolsai-usagebarQuota and reset times for 24 AI providersghpendingEvery open PR and issue in one listtclockA terminal clock with command widgetsFrank projectsExperiments that scratch my own itchesai-memoryLong-term memory for coding agentsai-jailAn OS sandbox for AI coding agentsWritingWritingThe blog and the LLM benchmarkWhy AGI doesn't matterAn opinion, with sourcesAI and studentsWhy foundations matter more than everThe limits of LLMsWhat AI can copy, and what it can'tCompetent Enough, the bookA year of putting AI to the testNewsletterThe M.Akita Chronicles, every MondayPodcastsLong conversations about AI, on videoGamesMy game collection as code, on OmarchySetupOmarchy, tmux, and what I set asideFollow
Book

Competent Enough

A year putting artificial intelligence to the test, without fear and without hype.

Competent Enough as a 3D paperback: a giant retro-futurist robot with a weary round face kneels at a 1950s World's Fair while a crowd watches. The title runs down the navy spine.

I took the year's articles from my blog, reorganized and rewrote them as one story, fact-checked them again and illustrated them. It's for anyone anxious or confused about AI: the professional afraid of being replaced, the parent who read that AGI is two years away, the manager being sold an "AI transformation".

Read it here

The whole book, page by page, without downloading anything. Full screen shows two pages side by side, and the reader picks up where you stopped.

Opening the book…

What's inside

Five parts, each answering a question people keep asking me about AI.

462
pages in the print edition
20
chapters
146
terms in the glossary
295
references
  1. What is this thing, anyway

    Is it intelligent? Is AGI coming?

  2. The marathon

    Does it actually work, or is it all demo?

  3. Numbers and propaganda

    Which AI is best? Should I believe the announcements?

  4. Who holds the keys

    Is it dangerous to let it touch my things?

  5. What about my job?

    Will it take my job? What comes next?

Plus two interludes, an optional appendix on how the machine works inside (no math), a timeline, practical checklists and the list of the original articles behind each chapter.

Making-of

How 131 blog posts, almost half a million words, became one book written with agents, without losing control of what it says.

  • Every post went into a sealed inventory with a hash, and a lineage map says which post superseded which, so the book tells the latest state of what I measured.
  • Every paragraph carries an invisible marker of the post it came from. When I edit a post, one command lists the paragraphs to review.
  • Every number, date and name was checked against the posts, the reference projects and primary sources. Some of the corrections were to my own blog.
  • Agents wrote the drafts. Reading, cutting, correcting and approving every chapter stayed with me.

The agents were competent enough to write the draft. Deciding what it said remained my job.

Os agentes foram competentes o suficiente pra escrever o rascunho. Decidir o que ele dizia continuou sendo trabalho meu.

Everything here is open source

The tools, the skills and the benchmark are public repositories. Fork them and build the version that fits the way you work.