AITypeSafe AI accused of presenting an earlier open-source architecture as a new invention
A user alleges that an independent Indian developer published and open-sourced the same single-pass reinforcement-learning design a year before the lab's announcement.
TLDR
In a September 19 post, a user alleges that TypeSafe AI announced a “breakthrough” System 1 architecture matching a design an independent Indian developer had built, published and open-sourced a year earlier. The user claims the lab generalized the architecture, gave it a new acronym and presented it as a novel invention.
Combined views
182.8K
2 Sources, first seen 1d ago
likes
TypeSafe AI accused of presenting an earlier open-source architecture as a new invention
A user alleges that an independent Indian developer published and open-sourced the same single-pass reinforcement-learning design a year before the lab's announcement.