Path 1 · 3 steps Draft
The job after AI
What product management became once building stopped being the expensive part, where the value went, and how to hire for the job as it is now.
0 of 3
Start →Bets & Ships — product newsletter for product leaders and hiring managers
Everything is repricing
Product roles, design roles, software pricing, and software valuations are all moving at once. It's one story, and most people are reading it four separate times.
Most popular
View all →From the channel
Bets & Ships on YouTube
The essays, read aloud and drawn, for when you would rather watch.
- Essays, narratedRead aloud, start to finish.
- Drawn scene by sceneEvery chart and idea, drawn.
- The essay behind itLinked, with every source.
Latest · Everything is repricing · 8 min · The 60-second version →
Recent posts
-
The JobResearch got cheap. Evidence didn’t.
AI made every step of product research faster: recruiting, interviewing, synthesizing, even the users. It didn’t make it cheaper to know whether any of it is true. What the evidence says, and how to do research now.
-
The JobYour judgment is fine. Your threshold may not be.
How to keep judging well when the volume of plausible artifacts goes up tenfold: what the detection research actually says, and what to do about it.
-
The JobThe feedback problem
New habits for product leaders in 2027: what to drop, what to build, and which part of the job each one fixes.
-
BetThe renewal cliff
A 2027 risk map for B2B SaaS: what’s actually dangerous, what only looks dangerous, and an 18-month plan.
-
The JobThe middle is the part being removed
Where humans actually sit in product management after AI — and why “orchestrator” is the wrong answer.
-
EvalsThe rubric that replaced the spec
You cannot write an acceptance criterion for a system that answers differently every time. What you can write is the standard you will judge it against.
Learning paths
All paths →Not complete yet. The Evals path is ready to follow; the other four are drafts, and more steps are coming.
Path 2 · 2 steps Draft
Judgment and feedback
Why reading AI output is not a check, and the few habits that tell you, soon and plainly, whether you were right.
0 of 2
Start →Path 3 · 5 steps
Evals for product people
From “I tried it and it seemed fine” to a score you can rerun. By the end you will have real cases, a rubric you wrote, and a first run to compare against.
0 of 5
Start →Path 4 · 2 steps Draft
Hands-on tools
The reading skills and the kit a product leader needs now: what to read in the code, and what to use, choose and test in each category.
0 of 2
Start →Path 5 · 2 steps Draft
The market and the money
One mechanism behind the headlines about roles, pricing and valuations, and a way to place your own product on the risk map.
0 of 2
Start →Series
All series →Reference
The 2026 product leader’s toolset
Every tool the job now runs on, what each one is actually for, and where it stops being worth the setup.
- 1The reasoning partner
- 2The prototyping tool
- 3The agentic coding tool
- 4Direct access to your data
- 5Research synthesis
- 6The eval harness
- 7The system of record
- 8The watcher
4 parts
The job, rewritten
What product management became once building stopped being the expensive part — one argument at a time.
- 1The product manager, before and after
- 2Evals: the product skill nobody taught you
- 3Should a PM know how to read code in 2026?
- 4How to hire a PM now
Case Studies
All case studies →-
Case study
Rubric — a PM-shaped eval tool
Bring 50 real cases and a rubric written in plain English, and get a score you can rerun after any change. The screen that matters is the diff: which cases moved, and in which direction.
-
Case study
Scorecard — a launch prediction ledger
Write down what success looks like before you ship, and keep the verdict whichever way it lands. Unclear counts as an answer, and the ledger says how good your predictions have actually been.
-
Case study
Synthesizer — themes from interview transcripts
Reads a set of transcripts and returns themes and contradictions, where every claim is built from quotes checked word for word against the transcript they came from. A quote it invents is caught on the page, not by you, later.
The Bets & Ships newsletter
Get the next one by email
For product leaders and hiring managers: how product work is changing after AI, with the evidence attached.
- One essay every TuesdayHow product work is changing.
- The evidence attachedCharts, numbers, every source.
- Case studies from the buildThe tools, and what broke.
Free. You’ll get one email to confirm first.