The AI circuit has a credibility problem: everyone "saw it coming," and nobody can show you where they said so. Predictions get made in vague language, on stages, with no dates — so they can never be wrong, which means they were never really predictions.
This page is the fix, applied to myself first. The Prediction Ledger is a standing record of the calls I have made publicly — each with its source, its specific claim, and a scheduled review where I grade it in public: right, wrong, or too early. Future entries will add new calls before they're fashionable, because that's the point of a ledger: you write it down when it's uncomfortable, not after it's obvious.
Why you should trust a scored record more than a confident stage
A track record with no misses is a track record with no bets. I have been early and right — committing to digital for the trades in 2006, to AI as the platform shift years before my industry would say the words — and I have paid the price of early often enough to respect how hard calling the future actually is. The essays in Move First tell those stories. This ledger simply makes the habit auditable going forward.
On the record as of July 2026
| # | The call | Where I said it | Review |
|---|---|---|---|
| 001 | Traditional websites are becoming obsolete. The brochure-site model — pages built to persuade a scrolling human — loses its primacy as AI assistants become the first point of contact; web presences must be rebuilt to feed machines (entities, structured data, machine-readable indexes), not just impress visitors. | The Kraft Your Life Radio Show, ep. 144 (May 2026), and stages before it | December 2027 |
| 002 | AI assistants and agents will reshape customer buying behavior in home services. Siri, Alexa, and the next generation of agents move from novelty to the front door of the transaction — answering, comparing, and increasingly booking on the customer's behalf. | Kraft Your Life, ep. 144; Toolbox for the Trades (ServiceTitan) | December 2027 |
| 003 | The winning web architecture is the highest-efficiency site: the lowest volume of code with the highest volume of content. Bloated page-builder stacks lose to lean, content-dense, machine-legible builds. | Kraft Your Life, ep. 144 | December 2027 |
| 004 | AI agents take over booking, dispatch, and customer follow-up work. The coordination layer of a home services business — not the wrench work — is where agents land first and hardest. | Kraft Your Life, ep. 144; Matt Chatts, ep. 53 | June 2028 |
| 005 | AI is the future of marketing for the trades — and the contractors who don't adapt will be left behind when the wave hits. Not a slogan: a competitive statement about who wins discovery, trust, and the sale as search behavior moves into AI systems. | HARDI webinar — "AI: The Future of Marketing"; Hands Up, ch. 12 | December 2028 |
The rules this ledger runs on
- Every entry has a source. A prediction only enters the ledger with a link to where it was made publicly — a stage, a show, a chapter, or this blog.
- Every entry gets reviewed on schedule. Reviews are published as ledger essays; grades are right, wrong, or too early — and "too early" only counts once per call.
- Misses stay on the page. Nothing gets quietly deleted. The misses are the proof the hits were bets.
The first annual scorecard publishes at the end of 2026. If you want the fuller version of where I think this is all going — the decade, not the quarter — chapter twelve of Hands Up is the long-form forecast, free like the rest of the book. And if your event wants the version with a microphone: that's here.