Foundational models are not enough to predict AI’s future, argues a skeptic

StewartalsopIII · x · 2026-07-26

The post argues that foundational models would only be indispensable if deterministic code did not exist, and if they were not already good at writing deterministic code. Since both conditions are false, the author says we are in “no man’s land” when trying to predict the future of AI.

It is a compact skepticism about how much predictive power current model-centric narratives really have, rather than a product or research announcement.

Original post →

More from AGI Musings

AGI Musings channel →