Product & Craft · Week of August 16, 2026
Cursor's platform gambit, DeepSeek's price reversal, and the verification economy reshape AI-native product economics
- Cursor's $100M-ARR sprint and the pivot from single product to multi-product platformArticle — "Cursor's $100M-ARR sprint and the pivot from beloved single product to multi-product platform": A user flew from Tokyo to San Francisco in August 2024 carrying the cover proof of the first book ever written ab…
- DeepSeek hikes API prices up to 1,100% with V4-Pro — is the subsidy era ending?Cache-hit input tokens on DeepSeek-V4-Pro just went from $0.003625 per million to $0.044 at peak. That's, by this article's reading, the 1,100% headline — a number that reads like the price war's obituary, cheap Chinese …
- The verification layer becomes the bottleneck: valuations, AI review, and eval toolingThe verification layer becomes the bottleneck: Blacksmith's 10x valuation, AI review outrunning humans, and eval tooling": Two engineers are looking at the same problem and drawing opposite conclusions about what your mo…
- Grok emerges as a credible third AI platform — with a data-access blind spot at its coreClaire Vo posted the tell: people she knows are "skipping the anthropic/openai debate entirely and becoming grok bois instead."
- Specs for two readers: how PMs are rewriting documentation for humans and AI agentsA PM writes a 400-line spec. She reads the first 50 lines and the last 20, then skims the rest.
Threads we are following
- Whether the rest of Cursor's bundle — collaboration and agent products — ships as fast as Composer 2.5 did
- Whether DeepSeek's actual V4-Pro increase stays under KB's predicted 'no more than double' ceiling once export-market rates settle
- Whether next round of eval benchmarks moves the 30% human-judgment recovery number, or whether capital keeps flowing to throughput alone
- Will xAI ship a working native X connector for Grok Bot — cashing the 'soon' promise — and turn its marketed moat into a shipped one?
- Whether spec templates like Obvious's mandatory scope/decided fields become a standard PM practice, or fragment into competing formats
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.