Two valid product strategies: bet on frontier models getting cheaper, or cheap models getting smarter

tobowers · x · 2026-09-23

Toby Shorin lays out two equally valid ways to think about current model capabilities: design your product around frontier models assuming they'll get cheaper, or design around cheap intelligence assuming it'll get better. In a follow-up he notes the two paths are subtly different — cheap currently equals fast.

Related event: Two product bets: frontier models getting cheaper vs cheap models getting smarter(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →