The GP's CTO: before betting on AI, ask "what if we're wrong?"
every · x · 2026-08-25
Dan Pup, CTO at The GP, proposes separating AI-bet assumptions into two kinds before committing: easily revisable ones (e.g., falling token costs just change product behavior) and structural ones (e.g., deep lock-in to one model provider where leaving means rewiring the product).
A wrong strategy can still be survivable; the dangerous bets are those the product cannot unwind. He suggests adding "What happens if we're wrong?" to every roadmap conversation.
More from Companies & People
- Lambda onboards a group of new affiliates to its official X organization — TheZachMueller · 2026-08-25
- Thomson Reuters accused of using unlicensed data via Chinese open models — BlancheMinerva · 2026-08-25
- South Korea's Sovereign AI Initiative Selects LG, SKT, and Upstage as Finalists — ArtificialAnlys · 2026-08-25
- Microsoft Montreal hosts "Inner Loop" exclusive ML seminar — jm_alexia · 2026-08-25
- Sam Altman: OpenAI would not have existed without Y Combinator — garrytan · 2026-08-25
- User praises OpenAI's transparency, criticizes Anthropic's silence — soumitrashukla9 · 2026-08-25