Tesla FSD v15 starts rolling out with claimed 25X safety-critical improvement and shared vision encoder for Robotaxi and Optimus
XFreeze · x · 2026-11-22
Tesla has begun rolling out FSD v15, claiming a 25X improvement in handling safety-critical situations via an upgraded vision encoder shared across FSD, Robotaxi and Optimus, trained on millions of internet videos.
Key changes:
- Tesla Reality View: a new real-time generative driving visualization showing what FSD plans to do, replacing separate model signals with one view while using less compute
- Smarter navigation: learns route preferences from the fleet, starting with trip starts and ends
- Personalized parking: learns your parking preferences and suggests multiple options
- Better detection: road signs, small animals, debris, chains, potholes, deep standing water; slows for speed cameras
- Misc: improved driver monitoring with hats/sunglasses, new Bike Rack Mode, FSD-based Autoshift predictions
The shared vision foundation signals Tesla is converging its vehicles and humanoid robots on a common world-understanding model.
Related event: Tesla Rolls Out FSD V15 With New Architecture and Reality View(37 posts)→
More from Embodied
- Dev demos hands-first interactions for Meta VR glasses in Meta VR Start competition — tristanbob · 2026-11-22
- Musk highlights Tesla FSD v15 claims: 25x improvement in safety-critical situations — elonmusk · 2026-11-22
- Musk Hails FSD v15 Navigating Around Huge Road Branch as 'GTA in Real Life' — elonmusk · 2026-11-22
- Musk amplifies Tesla FSD V15 feature letting drivers pick any parking spot — elonmusk · 2026-11-22
- Robotics veterans clash: general models are the key to the last 6% problem — BradPorter_ · 2026-11-22
- Google DeepMind and YC startup Hebbian host world's first Robot Use Agents hackathon — DynamicWebPaige · 2026-11-22