Tesla’s FSD Beta program has over 100,000 drivers actively participating, Elon Musk revealed during his TED interview. Musk also stated that Tesla would achieve full-self driving this year — this time, he is more confident.
Chris Anderson hosted Elon Musk during his latest TED talk interview. Anderson asked about Musk’s latest timeline for Tesla to achieve full self-driving. The TED host clarified that by “achieving full self-driving” he meant Tesla cars could drive around most cities without human interventions, safer than a human driver.
“Yes, I mean, the car currently drives me around Austin most of the time with no interventions,” Musk replied. And we have over 100,000 people in our full self-driving beta program.”
Anderson delved deeper into the development of Tesla’s self-driving program to better understand why Musk’s previous one-year or two-year timeline for the autonomous software didn’t reach fruition. Musk explained there were “so many false dawns” in the self-driving program. There were plenty of times when Tesla believed it was making actual progress. He said that Tesla’s progress was like a log curve.
Speaking about log curves, Elon Musk told Anderson: “It goes up — you know — sort of fairly straight away. And then it starts tailing off, and you start getting diminishing returns.”
Musk shared that Tesla’s journey with full self-driving was a series of log curves for a bit where it felt like they were making strides, but then progress seemed to hit a ceiling, and they’d have to start again. However, it appears that Tesla has gone down the right path now.
“And ultimately, These things — you know — in retrospect, they seem obvious, but in order to solve full self-driving properly, you actually have to solve real-world AI,” Musk said.
The realization that Tesla has to solve real-world AI led to adoption and further developments of Tesla Vision — the company’s camera-only strategy for autonomous driving. With Tesla Vision, the company’s cars function more like a human with eyes. So for Tesla, neural nets act as the brain, and the cameras act as eyes — a connection that has ultimately led to Tesla’s humanoid robot Optimus.
Elon Musk believes that with the current architecture, Tesla can increase the probability of vehicles avoiding an accident compared to a human driver behind the wheel. He attributed his confidence in the new architecture to Tesla reaching high-quality unified vector space for full self-driving.
Watch Elon Musk’s recent TED talk with Chris Anderson below!
The Teslarati team would appreciate hearing from you. If you have any tips, reach out to me at maria@teslarati.com or via Twitter @Writer_01001101.
News
Tesla is improving this critical feature in older vehicles

Tesla is set to improve a critical feature that has not been present in older vehicles with a new update.
Tesla vehicles feature a comprehensive suite of driver assistance features, some of which aid in driving itself, while others support the vehicle’s surroundings.
One of those features is that of Driver Visualization, and with the rollout of a new update, owners of Intel-based Tesla vehicles are receiving an upgrade that will come with a simple software update.
Tesla plans to use Unreal Engine for driver visualization with crazy upgrade
The update will provide new visualizations while Intel-based vehicles are in reverse, a feature that was not previously available, but will be with Software Update 2025.32.2.
The improvement was spotted by Not a Tesla App via TheBeatYT_evil:
Noticed something new in 2025.32.2 on my Intel MCU + USS car with FSD.
When shifting into reverse, the full FSD visualization now stays on instead of switching to the old plain autopilot visuals.
Might be small, but it makes backing up feel more seamless. pic.twitter.com/o44levkdtM
— Beat (@TheBeatYT_evil) September 5, 2025
Previously, vehicles Tesla built were equipped with Intel-based processors, but newer cars feature the AMD chip, which is capable of rendering these visualizations as they happen. They were capable of visualizations when driving forward, but not in reverse, which is what this change resolves.
It is a good sign for those with Intel-based vehicles, as Tesla seems to be paying attention to what those cars are not capable of and improving them.
This was an undocumented improvement associated with this particular update, so you will not find any mention of it in the release notes that Tesla distributes with each update.
News
Tesla looks to make a big splash with Robotaxi in a new market
Tesla has been transparent that it is prioritizing safety, but it believes it can expand to basically any geographical location within the United States and find success with its Robotaxi suite. CEO Elon Musk said it could be available to half of the U.S. population by the end of the year.

Tesla is looking to make a big splash with Robotaxi in a new market, as the company was spotted testing validation vehicles in one region where it has not yet launched its ride-hailing service.
After launching Robotaxi in Austin in late June, Tesla followed up with a relatively quick expansion to the Bay Area of California. Both service areas are operating with a geofence that is expansive: In Texas, it is 173 square miles, while in the Bay Area, it is roughly 400 square miles.
Tesla has been transparent that it is prioritizing safety, but it believes it can expand to basically any geographical location within the United States and find success with its Robotaxi suite. CEO Elon Musk said it could be available to half of the U.S. population by the end of the year.
There have been plenty of reports out there that have speculated as to where Tesla would land next to test Robotaxi, and Nevada, Florida, Arizona, and New York have all been in the realm of possibility. These regions will need to approve Tesla for regulatory purposes before Robotaxi can officially operate.
Tesla is still testing and performing validation in several regions, and in Tempe, Arizona, things are moving forward as a Model Y with a LiDAR rig was spotted performing ground truth for the platform:
🚨 BREAKING: Just caught Tesla Robotaxi test vehicles cruising in Tempe, AZ! Rollout coming soon! pic.twitter.com/Oanw0Zx5pP
— Adub08 (@adub0808) September 10, 2025
With the LiDAR unit, many followers of the self-driving and autonomy space might wonder why Tesla uses these apparatuses during validation, especially considering the company’s stance and vision-based approach.
LiDAR is used for “ground truth,” which is basically a solidification or confirmation of what the cameras on the car are seeing. It is a great way to essentially confirm the accuracy of the vision-based suite, and will not be used on Robotaxi units used within the ride-hailing suite.
The Robotaxi platform was made available to the public earlier this month, as Tesla launched its app for iOS users.
Tesla Robotaxi app download rate demolishes Uber, Waymo all-time highs
Downloading the app allows you to join a waitlist, giving you the opportunity to utilize and test the Robotaxi platform in either Austin or the Bay Area.
News
Tesla hacker finds lifesaving FSD suggestions in 2025.32.3
The feature could drastically reduce instances of drivers operating their vehicles while distracted or exhausted.

A Tesla hacker has shared references to what could very well be lifesaving FSD suggestions in software update 2025.32.3.
The feature could drastically reduce instances of drivers operating their vehicles while distracted or exhausted.
New FSD features
As per longtime Tesla hacker @greentheonly, the EV maker’s drowsiness and lane departure suggestions in 2025.32.3 reference FSD. The hacker shared two alerts that specifically suggest the use of FSD, such as “Lane drift detected. Let FSD assist so you can stay focused,” and “Drowsiness detected. Stay focused with FSD.”
The hacker noted that the updated messages in 2025.32.3 are quite interesting because Tesla still advises drivers to be fully alert when using FSD Supervised. Thus, it is quite interesting to see the company advising users to use FSD when they seem unfocused or tired. That being said, FSD is still a supervised solution for now, but the system itself is already very capable, so the company’s updated alerts are not surprising.
Steps to FSD Unsupervised
Such warning messages would definitely make more sense once Tesla actually rolls out FSD Unsupervised. The system seems to be just waiting for regulatory approval for now, considering that it is already being used in vehicles that are coming off the production line at the Fremont Factory and Giga Texas. Tesla is also now using a driverless system for its Robotaxi services in Austin, as well as its autonomous ride-hailing service in the Bay Area.
For now, Tesla is hard at work preparing for the impending rollout of FSD V14, which Elon Musk has stated will be a notable improvement from the already-impressive performance of FSD V13. As per Musk, Teslas running FSD V14 will feel “sentient.” He also noted that the system “feels alive.”
-
News2 weeks ago
Tesla is overhauling its Full Self-Driving subscription for easier access
-
Elon Musk2 weeks ago
Elon Musk shares unbelievable Starship Flight 10 landing feat
-
Elon Musk2 weeks ago
Elon Musk reveals when SpaceX will perform first-ever Starship catch
-
Elon Musk1 week ago
Tesla’s next-gen Optimus prototype with Grok revealed
-
Elon Musk2 weeks ago
SpaceX Starship Flight 10 was so successful, it’s breaking the anti-Musk narrative
-
News5 days ago
Tesla launches new Supercharger program that business owners will love
-
Elon Musk5 days ago
Tesla Board takes firm stance on Elon Musk’s political involvement in pay package proxy
-
News1 week ago
Tesla appears to be mulling a Cyber SUV design