Product & Craft · Week of July 12, 2026
AI's coding-model glut forces new tooling, pricing, and PM playbook decisions
- Four coding models drop in one week: how builders now stitch models instead of picking oneHere's a contradiction worth sitting with. Claire Vo ran five frontier models through her own benchmark and declared GPT-5.6 Sol the winner "by a meaningful margin." In the same breath, she named a different model for ne…
- It's not the model, it's the harness: orchestration and DIY benchmarks become the real skillHere is the contradiction most builders haven't noticed yet: the same person who tells you "it's not the model, it's the harness" is also the person who just discovered her own benchmark disagrees with her taste almost c…
- SaaS builds AI features but can't charge for them: why PMs must now own pricingStart with the contradiction that should keep you up at night.
- Usage looks healthy, the customer churns: new frameworks to measure outcomes over engagementHere is a contradiction worth sitting with. Sturdy will tell you that a customer with high usage might be your biggest churn risk — opening support tickets, emailing their CSM, hammering the product not because they love…
- The PM playbook rewrites itself: leaner generalist pods, blended roles, outcome roadmapsThe PM playbook rewrites itself: leaner generalist pods, blended roles, outcome roadmaps
Threads we are following
- Will product and eng orgs create an actual budget line and owner for the standing re-evaluation tax, or keep absorbing it invisibly?
- Whether anyone builds an automated verifier that can catch a broken prototype on the first screenshot the way a human eye does
- Whether Microsoft's Copilot repricing moves the paid weekly-active number above 1% next quarter
- Whether Sharma's Outcome Clarity Test, Tidwell's Retention Architecture, and Obasuyi's Duplo metrics converge into a shared standard, or fragment into competing frameworks
- Whether 'product staff' as a formal title spreads beyond Meta-scale orgs, or stays a one-company experiment
One issue a week. Free.
sift. reads hundreds of sources so you don't have to, and sends one issue you can actually finish. No daily noise. Unsubscribe anytime.