Jitendra Malik: Stop Conflating VLMs with World Models in Robotics
JitendraMalikCV · x · 2026-08-24
Berkeley professor Jitendra Malik argues for terminological precision in robotics: VLM, VLA, and World Models are being used indiscriminately. VLMs arose as multimodal extensions of LLMs trained on tasks like VQA, capturing static scene semantics with no dynamics. World Models (LeCun, Ha & Schmidhuber 2018) are primarily dynamics models rooted in control theory (Bellman, Kalman, 1960)—classically written a priori from physics, today learned from video and robot trajectories, but the concept of state-transition modeling is unchanged and shouldn't be mixed with VLMs.
More from Embodied
- Predicting AGI and ASI by Year-End, Including Self-Driving Cars and Humanoids — davidpattersonx · 2026-08-24
- Humanoid robots enter mass production as human fertility drops below replacement — Dr_Singularity · 2026-08-24
- Galbot humanoid robot completes 100+ consecutive tennis rallies autonomously — Distinct-Question-16 · 2026-08-24
- DaxAI's rideable robot horse Qiji X1 debuts with 1400 Nm joint motors — tristanbob · 2026-08-24
- Memory Legend Predicts Widespread ANN Use in Edge IoT and Humanoids — BenBajarin · 2026-08-24
- Spent $266 on 4 Local Models to Unlock Amazon Tablet — yogthos · 2026-08-23