3D AI research beyond incremental gains
A user says they've watched stronger foundation models rapidly absorb entire research directions in 3D, and sees bigger questions in scaling, data and model limits.
TLDR
One user describes seeing entire research directions in 3D absorbed by stronger foundation models surprisingly quickly. In their view, the interesting question is shifting from “what module gets another +0.2%” to what actually scales, which data creates new capabilities and where models fundamentally break.
Combined views
31
1 Source, first seen 2h ago
2 reposts
3D AI research beyond incremental gains
A user says they've watched stronger foundation models rapidly absorb entire research directions in 3D, and sees bigger questions in scaling, data and model limits.