·AI
The Voice Is the Joke
I rebuilt Knora's English cast, deleted the style axis, and found two audio defects every test said were fine. The update is live — and the lesson is not really about comedy.
#Build in Public
11 posts with this tag
·AI
I Built an App With Almost No UI. That Was the Point.
The best review of my comedy app came from a table with four friends and a round of drinks — and my first version failed it. So I removed the joke format, then I removed the interface. Here is what a one-gesture app forces, why a 75-second ceiling is a table-length constraint, and the uncomfortable part for anyone whose day job is attention metrics.
·AI
I Gave My 9-Agent Marketing Platform Away as Open Source
STRAŦUM v1 is now MIT-licensed and public. Nine agents, two Postgres schemas, ten locales — and a table that shipped with Row Level Security switched off, reachable by anyone holding the key that loads in the browser. Here is what I gave away, what I would not build the same way again, and why you are welcome to fork it and build your own.
·AI
My AI Bill Went From $200/Month to $45. I Shipped More, Not Less.
June: $57.01. July: $33.09. Two months after switching to DeepSeek V4 Pro via OpenCode, my AI costs dropped 75%. The commit history proves I shipped more, not less — 786 commits across 4 products. Here's the full breakdown.
·AI
Offline Mode for My AI App Didn't Need AI. It Needed Boring.
I almost built a clever, AI-specific offline mode for DIALØGUE's iOS app. Then I asked a more boring question — what does iOS already give me? — and the real work turned out to be resisting the urge to be clever. Here's the standard, unglamorous machinery behind "download an episode and listen on a plane," and why boring was the senior choice.
·AI
Native Isn't a Skin: Rebuilding DIALØGUE's iOS App
I shipped DIALØGUE's iOS app as a port of the web product, then rebuilt it natively — three tabs, lock-screen audio, a synced transcript, resilient offline, and Siri — because a web app shrunk to a phone is still a web app.
·AI
I Rolled Back 2,724 Lines After One AI Audio Change Broke Production
I pushed what should have been an upgrade to my podcast platform's voice system. Six days and several commits later, I deleted 2,724 lines of code and rolled back to what worked. Here is what happened and what it taught me about testing production AI changes.
·AI
I'm Using Claude Code for Everything Else But Coding
Nineteen days after cancelling Claude Max, my pattern has settled. Codex with GPT-5.4 on xHigh took the coding seat. Claude Code with Opus 4.7 on xHigh took every other seat at the desk.
·AI
Two years after my 7 Andrew Ng courses: the 2026 path I'd actually take
People still message me asking if the 7 Andrew Ng courses I recommended in 2023 are the right path today. Short answer: most of them, yes, but with a different roadmap around them. Here is my 2026 update, with per-course verdicts and a forked path for builders and operators.
·AI
Why I Cancelled Claude Max After 13 Months and What I’m Testing with Codex Next
I cancelled Claude Max after 13 months and US$1,892.38 in subscription fees. This is not a victory lap. It is a 30-day test to see whether I can keep shipping STRATUM, DIALOGUE, my course platform, and this site at the same pace with Codex as my primary tool.
·AI
I'm Dropping My $200 Claude Code Plan After Two Weeks with Codex
Two weeks after my original comparison, both tools shipped major updates. Codex challenged my product strategy in ways Claude Code did not. Claude Code shipped Agent Teams and AutoMemory. The result: I am cutting my $200/month Max plan — and getting better output for less money.