Tesla’s AI chief, Ashok Elluswamy, has revealed that Full Self-Driving (FSD) should learn how fast each owner prefers to travel, instead of relying on a manual ceiling. This approach prioritizes advanced FSD capabilities over user intervention, with Elluswamy calling maximum-speed control “an anti-pattern.”
Elluswamy Pushes FSD Toward Learned Preferences
Elluswamy stated that the company is working on better learning of users’ implied preferences, responding to Tesla enthusiast David Moss who hoped that Max Speed control would never return on either AI3 or AI4 vehicles. Elluswamy’s comment suggests a more personalized FSD that studies interventions, driving profiles, traffic flow, and road context.
“Max speed control is an anti pattern. We are working on better learning of user’s implied preferences.” - Ashok Elluswamy
Tesla’s FSD Development
Tesla’s documentation indicates that FSD (Supervised) selects an appropriate maximum speed using the detected limit, driving profile, and surrounding traffic. Depending on the software version and market, owners can adjust speed with the steering-wheel scroll control or select profiles ranging from Chill to Mad Max. However, Tesla still requires drivers to remain attentive and responsible for vehicle speed.
Owners Split Over Removing Manual Controls
Supporters of Elluswamy’s AI-first approach argue that capable autonomy should not need repeated manual correction. However, other owners have pushed back, saying that FSD sometimes travels below the limit or accelerates beyond their comfort level. The divide among owners highlights the challenges of developing a personalized FSD system.
European Regulators Scrutinize Tesla Speed Features
The debate over FSD’s speed features has also caught the attention of European regulators. Sweden’s transport authority has urged European regulators to oppose a wider FSD rollout unless Tesla disables behavior that permits speeds above legal limits. France has also raised concerns about speeding and driver attention, adding pressure to Tesla’s European FSD expansion.
Future Developments
Tesla is separately developing FSD that remembers parking preferences and accepts voice commands for neighborhood navigation. The strategy aims to replace menus and overrides with software that predicts intent. As things stand, FSD remains a Level 2 driver-assistance system, requiring the driver to stay fully engaged.
Market Impact
Tesla’s stock scores poorly on Momentum and Value metrics but provides satisfactory Growth and Quality. The company’s shares were up 0.60% to $324.02 during pre-market trading on Tuesday.
Conclusion
Tesla’s prioritization of advanced FSD capabilities over user intervention marks a significant shift in the company’s approach to autonomous driving. As the technology continues to evolve, it will be interesting to see how Tesla balances the need for personalized FSD with the concerns of regulators and owners.