aspiringforintelligence.substack.com
|
ksl
|
|
Two Madrona partners argue that frontier models from Anthropic, OpenAI, and Google will function like infrastructure – not as finished products – leaving massive room for startups at the application layer. They draw parallels to Ford’s assembly line and the Green Revolution, where automation destroyed specific jobs but generated far more economic activity than it displaced. The sharpest point is about the “blank slate problem”: raw model intelligence means nothing if users don’t know what to do with it, and that gap is where product taste beats raw capability. It’s a VC-flavored take, but the underlying tension is real – Amazon and Block have already cited AI efficiencies when cutting staff, even as the application layer remains wide open.
