🧭 THIS WEEK AT BuildProven
Howdy, a bit different this week, tried to build a ‘field guide’ to catch up where we’ve been and are now. Hope you like it.
as usual, would love it if you could share this newsletter, give feedback via the poll at the end, or best, just hit 'reply' and tell me what you need/want with AI. I will definitely reply and try to help best i can.🧰 Worth Your Click
Here are a few things I found recently:
more next week when out of vacay mode.
🗺️ FEATURED INSIGHT
McKinsey surveyed 1,993 organisations across 105 countries last year. 88% reported using AI somewhere in the business. About 5.5% reported that more than 5% of their profit could be attributed to it.
That gap is the whole story of AI at work right now. Adoption is universal. Value is rare. And the people closing it aren't the ones who read the most AI newsletters — they're the ones who understand enough of the machinery to point it at a real problem and check its work.
So this week I did something different. Instead of one build, I wrote the whole map: the page I wish someone had handed me at 52, when I left 25 years of automotive program management and started building with this stuff.
It's one page, and it's long on purpose. Five parts:
How we got here — the four breakthroughs that actually explain why the tools behave the way they do (1956, 1997, 2012, 2017). No more history than you need.
The vocabulary — what the words actually mean. Tokens, context window, hallucination, agent, RAG, MCP. The decoder, ordered so each term builds on the last.
Where AI earns its keep — the six categories where it reliably pays at work, and the one property they all share (you can verify the output). Plus where it fails, so you don't learn that the expensive way.
The tools people actually use — three tiers, and why the real gains live in the tier almost no non-developer has found yet.
Building something yourself — the part nobody tells professionals they can do. It ends with a prompt you paste tonight to make a folder of files you can never search suddenly searchable.
Read it once, then keep it open in a tab and come back to the section you need. That's what it's for.
The last section is the one that matters most. It's a real, copy-paste prompt that points Claude Code at a folder you own and builds you an index.md — a one-line summary, date, and named people for every document, newest first. Under a minute of its time. You'll learn more about what this technology is and isn't in that twenty minutes than in six months of reading about it.
Reply and tell me which folder you point it at, and what it gets wrong. The wrong answers are the interesting ones, and the good ones become a future issue's worked example.
Weekly build logs from a 25-year program manager who codes with AI.
— Brett
👉 Hit “Reply” and share your experience — I read every one!
Picture by xxx on Unsplash.
