Walter Isaacson’s Elon Musk biography is set to be published on Tuesday, and a new preview of the book illustrates details about Tesla’s development of the upcoming Full Self-Driving (FSD) version 12.
In an additional preview of his Musk biography for CNBC, Isaacson discusses the use of AI in the development of Tesla’s FSD v12, in a shift that took place within the last several months. Isaacson talks about Tesla’s recent development of the upcoming FSD v12, which he and Tesla demonstrate has moved away from a “rules-based” approach.
Notably, FSD v12 is expected to use billions of video frames from real-world driving incidents to train its neural network rather than using thousands of lines of code like previous versions. In a conversation with Musk last December, Tesla Autopilot employee Dhaval Shroff had likened the concept to the popular chatbot ChatGPT, instead for use with driving.
“It’s like ChatGPT, but for cars,” Shroff said. “We process an enormous amount of data on how real human drivers acted in a complex driving situation, and then we train a computer’s neural network to mimic that.”
Surprisingly enough, Tesla only shifted toward this “neural network planner” approach recently. By the beginning of this year, however, the neural network had already analyzed 10 million video clips based on the best-case-scenario drivers the system had access to. Musk instructed employees at the company’s Buffalo, New York facility who were in charge of analyzing the footage to train the AI on things “a five-star Uber driver would do.”
Moving from a rules-based to a network-path-based AI approach allowed FSD to use human driving data to avoid obstacles, even if breaking some rules was necessary. Shroff helped demonstrate the idea to Musk with a demo featuring trash bins, debris, and upturned traffic cones, which the car handled surprisingly well.
“Here’s what happens when we move from rules-based to network-path-based,” Shroff explained. “The car will never get into a collision if you turn this thing on, even in unstructured environments.”
Musk quickly took to the idea, as can be seen in a recent livestream of Tesla’s FSD v12 software in Palo Alto with Autopilot software director Ashok Elluswamy. He has repeatedly spoken about the upcoming software version’s impressive driving results, despite one small moment in the drive where the car almost ran a red light.
In any case, Musk could argue that the red-light moment is a good case for the need for self-driving software to continually learn. Given that it will constantly be trained from the video data generated by camera footage from real-world drivers, it should theoretically make it safer over time, according to Musk.
During development, Musk also reportedly latched onto the fact that it took over a million video clips for the neural network to begin performing well, though he looks forward to what significantly more data will do for FSD.
Still, critics and regulators have expressed concerns about the faults of human drivers training AI-based driving systems, and Tesla has repeatedly been questioned by the National Highway Traffic Safety Administration (NHTSA) about its Autopilot and FSD beta systems.
According to Isaacson, Tesla plans to release FSD v12 as soon as regulators approve it. Meanwhile, an ongoing study by the National Highway Safety Board is looking to determine if self-driving cars should be permitted to imitate human driving actions that blur traffic rules, such as creeping up at stop signs.
Musk said in April that he expects Tesla to reach full autonomy within a year, though he has also been known to share ambitious targets for the software in the past.
You can read Walter Isaacson’s full account of the development of Tesla FSD v12 here, in a CNBC preview of the upcoming Elon Musk biography.
What are your thoughts? Let me know at zach@teslarati.com, find me on X at @zacharyvisconti, or send your tips to us at tips@teslarati.com.
News
Rivian unveils self-driving chip and autonomy plans to compete with Tesla
Rivian, a mainstay in the world of electric vehicle startups, said it plans to roll out an Autonomy+ subscription and one-time purchase program, priced at $49.99 per month and $2,500 up front, respectively, for access to its self-driving suite.
Rivian unveiled its self-driving chip and autonomy plans to compete with Tesla and others at its AI and Autonomy Day on Thursday in Palo Alto, California.
Rivian, a mainstay in the world of electric vehicle startups, said it plans to roll out an Autonomy+ subscription and one-time purchase program, priced at $49.99 per month and $2,500 up front, respectively, for access to its self-driving suite.
CEO RJ Scaringe said it will learn and become more confident and robust as more miles are driven and it gathers more data. This is what Tesla uses through a neural network, as it uses deep learning to improve with every mile traveled.
He said:
“I couldn’t be more excited for the work our teams are driving in autonomy and AI. Our updated hardware platform, which includes our in-house 1600 sparse TOPS inference chip, will enable us to achieve dramatic progress in self-driving to ultimately deliver on our goal of delivering L4. This represents an inflection point for the ownership experience – ultimately being able to give customers their time back when in the car.”
At first, Rivian plans to offer the service to personally-owned vehicles, and not operate as a ride-hailing service. However, ride-sharing is in the plans for the future, he said:
“While our initial focus will be on personally owned vehicles, which today represent a vast majority of the miles to the United States, this also enables us to pursue opportunities in the rideshare space.”
The Hardware
Rivian is not using a vision-only approach as Tesla does, and instead will rely on 11 cameras, five radar sensors, and a single LiDAR that will face forward.
It is also developing a chip in-house, which will be manufactured by TSMC, a supplier of Tesla’s as well. The chip will be known as RAP1 and will be about 50 times as powerful as the chip that is currently in Rivian vehicles. It will also do more than 800 trillion calculations every second.
Meet the Rivian Autonomy Processor.
Fast, smart, scalable and purpose-built for autonomous driving and the world of physical AI. Hitting the open road in 2026. pic.twitter.com/0wYXi5WKy7
— Rivian (@Rivian) December 11, 2025
RAP1 powers the Autonomy Compute Module 3, known as ACM3, which is Rivian’s third-generation autonomy computer.
ACM3 specs include:
- 1600 sparse INT8 TOPS (Trillion Operations Per Second).
- The processing power of 5 billion pixels per second.
- RAP1 features RivLink, a low-latency interconnect technology allowing chips to be connected to multiply processing power, making it inherently extensible.
- RAP1 is enabled by an in-house developed AI compiler and platform software
As far as LiDAR, Rivian plans to use it in forthcoming R2 cars to enable SAE Level 4 automated driving, which would allow people to sit in the back and, according to the agency’s ratings, “will not require you to take over driving.”
More Details
Rivian said it will also roll out advancements to the second-generation R1 vehicles in the near term with the addition of UHF, or Universal Hands-Free, which will be available on over 3.5 million miles of roadway in the U.S. and Canada.
More than any other feature, our owners have asked for more hands-free miles.
With Universal Hands-Free, you can now enjoy hands-free assisted driving on any road with clearly defined lanes. That’s roughly 3.5 million miles in the U.S. and Canada.
Look for it in our next… pic.twitter.com/ZFhwVzvt6b
— Rivian (@Rivian) December 11, 2025
Rivian will now join the competitive ranks with Tesla, Waymo, Zoox, and others, who are all in the race for autonomy.
News
Tesla partners with Lemonade for new insurance program
Tesla recently was offered “almost free” coverage for Full Self-Driving by Lemonade’s Shai Wininger, President and Co-founder, who said it would be “happy to explore insuring Tesla FSD miles for (almost) free.”
Tesla owners in California, Oregon, and Arizona can now use Lemonade Insurance, the firm that recently said it could cover Full Self-Driving miles for “almost free.”
Lemonade, which offered the new service through its app, has three distinct advantages, it says:
- Direct Connection for no telematics device needed
- Better customer service
- Smarter pricing
The company is known for offering unique, fee-based insurance rates through AI, and instead of keeping unclaimed premiums, it offers coverage through a flat free upfront. The leftover funds are donated to charities by its policyholders.
On Thursday, it announced that cars in three states would be able to be connected directly to the car through its smartphone app, enabling easier access to insurance factors through telematics:
Lemonade customers who own @Tesla vehicles in California, Oregon, and Arizona can now connect their cars directly to the Lemonade app! ⚡🚘
Direct connection = no telematics device needed 📵
Better customer experience 💃
Smarter pricing with Lemonade 🧠This is a game-changer… pic.twitter.com/jbabxZWT4t
— Lemonade (@Lemonade_Inc) December 11, 2025
Tesla recently was offered “almost free” coverage for Full Self-Driving by Lemonade’s Shai Wininger, President and Co-founder, who said it would be “happy to explore insuring Tesla FSD miles for (almost) free.”
The strategy would be one of the most unique, as it would provide Tesla drivers with stable, accurate, and consistent insurance rates, while also incentivizing owners to utilize Full Self-Driving for their travel miles.
Tesla Full Self-Driving gets an offer to be insured for ‘almost free’
This would make FSD more cost-effective for owners and contribute to the company’s data collection efforts.
Data also backs Tesla Full Self-Driving’s advantages as a safety net for drivers. Recent figures indicate it was nine times less likely to be in an accident compared to the national average, registering an accident every 6.36 million miles. The NHTSA says a crash occurs approximately every 702,000 miles.
Tesla also offers its own in-house insurance program, which is currently offered in twelve states so far. The company is attempting to enter more areas of the U.S., with recent filings indicating the company wants to enter Florida and offer insurance to drivers in that state.
Tesla Model Y configurations are getting hefty discounts and more benefits as the company is in the phase of its final sales push for the year.
Tesla is offering up to $1,500 off new Model Y Standard trims that are available in inventory in the United States. Additionally, Tesla is giving up to $2,000 off the Premium trims of the Model Y. There is also one free upgrade included, such as a paint color or interior color, at no additional charge.
NEWS: Tesla is now offering discounts of up to $1,500 off new Model Y Standard vehicles in U.S. inventory. Discounts of up to $2,000 are also being offered on Model Y Premiums.
These discounts are in addition to the one free upgrade you get (such as Diamond Black paint) on… pic.twitter.com/L0RMtjmtK0
— Sawyer Merritt (@SawyerMerritt) December 10, 2025
Tesla is hoping to bolster a relatively strong performance through the first three quarters of the year, with over 1.2 million cars delivered through the first three quarters.
This is about four percent under what the company reported through the same time period last year, as it was about 75,000 vehicles ahead in 2024.
However, Q3 was the company’s best quarterly performance of all time, and it surged because of the loss of the $7,500 EV tax credit, which was eliminated in September. The imminent removal of the credit led to many buyers flocking to Tesla showrooms to take advantage of the discount, which led to a strong quarter for the company.
2024 was the first year in the 2020s when Tesla did not experience a year-over-year delivery growth, as it saw a 1 percent slide from 2023. The previous years saw huge growth, with the biggest coming from 2020 to 2021, when Tesla had an 87 percent delivery growth.
This year, it is expected to be a second consecutive slide, with a drop of potentially 8 percent, if it manages to deliver 1.65 million cars, which is where Grok projects the automaker to end up.
Tesla will likely return to its annual growth rate in the coming years, but the focus is becoming less about delivery figures and more about autonomy, a major contributor to the company’s valuation. As AI continues to become more refined, Tesla will apply these principles to its Full Self-Driving efforts, as well as the Optimus humanoid robot project.
Will Tesla thrive without the EV tax credit? Five reasons why they might
These discounts should help incentivize some buyers to pull the trigger on a vehicle before the year ends. It will also be interesting to see if the adjusted EV tax credit rules, which allowed deliveries to occur after the September 30 cutoff date, along with these discounts, will have a positive impact.
Rivian unveils self-driving chip and autonomy plans to compete with Tesla
Tesla partners with Lemonade for new insurance program
