Tesla CEO Elon Musk has set a high bar for the next major leap in Full Self-Driving technology, claiming the upcoming version 15 will reach a level of safety that far exceeds human capabilities.
Responding to reviews of the current FSD v14.3 (which was supposed to be the last big puzzle piece), Musk noted that while minor point releases will continue to polish the current experience, the real breakthrough lies with v15. According to Musk, this version is designed to handle completely unsupervised and complex situations, which remains the ultimate goal for the automaker’s autonomous driving software.
The buzz around v15 centres on what Musk calls the large model. In the world of AI, parameters are essentially the connections or neurons in a digital brain. A model with more parameters can process more information and understand increasingly complex patterns. Musk confirmed that v15 will utilize a 10x parameter model, providing the car’s AI with a much higher capacity to learn from the trillions of miles of video data Tesla has collected.
This massive increase in brain power is expected to help the vehicle better predict unpredictable real-world scenarios, such as a cyclist swerving or a pedestrian hesitating at a busy intersection. Musk admitted that advancement with the current small model has been so quick that the larger architecture hasn’t quite caught up yet, but once it does, v15 will become the new flagship for the fleet.
For now, Tesla owners can expect polish updates in the v14.3 series to smooth out lane changes and fix minor bugs. However, the shift to v15 represents a massive architectural change that Musk frames as the version that finally makes true, unsupervised driving a reality.
Now, the bigger question is whether HW4 owners will be able to support FSD V15, or if it will require newer HW5 hardware. Many HW4 Tesla owners fear they might get left behind like HW3 vehicles, which have been waiting patiently still for FSD 14 lite, expected this summer.
Want to see more of our stories on Google?
P.S. — Buying a new Tesla? Click here to save $1,000 USD, while supporting independent news.
Help support us by shopping on Amazon here.
Links in this post are affiliate links, so we earn a tiny commission at no charge to you. Thanks for supporting independent media!
[…] Dutch vehicle authority RDW has granted type approval for Tesla’s Full Self-Driving (FSD) Supervised system, making the Netherlands the first European country to greenlight the advanced driver-assist […]
To the Tesla FSD Engineering Team,
I am a user on a 2019 Model X Raven (HW3), and I would like to share structured feedback alongside concrete, implementable proposals that I believe could significantly improve the FSD experience for legacy hardware owners.
—
THE PROBLEM
On HW3 vehicles running FSD v12 and earlier, the acceleration and deceleration logic follows the same aggressive profile as standard Autopilot. In urban environments (40–80 km/h), the car frequently drops abruptly from 60 km/h to 40 km/h and immediately accelerates back, creating a ride quality that is:
– Uncomfortable for all passengers
– Causing motion sickness in rideshare contexts (Uber drivers report passenger complaints)
– Not representative of how a human chauffeur would drive
This is not a minor inconvenience — it is actively damaging the perception of FSD among first-time passengers who experience it in rideshare vehicles.
—
PROPOSED SOLUTIONS
1. FSD-specific torque profile
Implement a dedicated torque cap when FSD is active, independent of the user’s Chill/Sport preference. A limit of approximately 150–200 Nm for urban speeds (under 100 km/h) would produce dramatically smoother acceleration. This is purely a software change — the hardware already supports it.
2. Speed chase dampening algorithm
Introduce hysteresis logic for speed limit changes in the 40–80 km/h urban range. If the detected speed limit fluctuates within a short distance or time window, FSD should interpolate smoothly rather than react immediately. This would eliminate the hard-brake sensation caused by sudden map or vision-based speed limit jumps.
3. Decouple FSD comfort from user drive mode
The Chill mode setting currently affects both manual driving and FSD behavior. A separate ‘FSD Comfort Mode’ toggle would allow users who enjoy Sport mode manually to still benefit from smooth autonomous behavior — and would serve as a sensible default for HW3 vehicles.
—
CONTEXT: WHY THIS MATTERS COMMERCIALLY
With FSD now approved in the Netherlands and a broader European rollout planned for summer 2026, European road conditions and passenger expectations will amplify these issues. European urban environments have more frequent and tighter speed limit transitions than US highways — exactly the scenario where this problem is worst.
Additionally, Uber fleet drivers using Tesla FSD are generating organic first impressions of the technology among thousands of passengers. Motion sickness complaints in that context represent reputational risk that is measurable and addressable.
—
Thank you for your continued work on FSD. I believe these proposals are low-effort, high-impact software changes that would meaningfully improve the experience for the entire legacy HW3 community.
Best regards,
Mr. Legacy
2019 Model X Raven | HW3 | FSD