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
Lucid unveils Lunar Robotaxi in bid to challenge Tesla’s Cybercab in the autonomous ride hailing race
Lucid’s Lunar robotaxi is gunning for Tesla’s Cybercab in the autonomous ride hailing race
Lucid Group pulled back the curtain on its purpose-built autonomous robotaxi platform dubbed the Lunar Concept. Announced at its New York investor day event, Lunar is arguably the company’s most ambitious concept yet, and a direct line of sight toward the autonomous ride haling market that Tesla looks to control.

At Lucid Investor Day 2026, the company introduced Lunar, a purpose-built robotaxi concept based on the Midsize platform.
A comparison to Tesla’s Cybercab is unavoidable. The concept of a Tesla robotaxi was first introduced by Elon Musk back in April 2019 during an event dubbed “Autonomy Day,” where he envisioned a network of self-driving Tesla vehicles transporting passengers while not in use by their owners. That vision took another major step in October 2024 when, Musk unveiled the Cybercab at the Tesla “We, Robot” event held at Warner Bros. Studios in Burbank, California, where 20 concept Cybercabs autonomously drove around the studio lot giving rides to attendees.
Fast forward to today, and Tesla’s ambitions are finally materializing, but not without friction. As we recently reported, the Cybercab is being spotted with increasing frequency on public roads and across the grounds of Gigafactory Texas, suggesting that the company’s road testing and validation program is ramping meaningfully ahead of mass production. Tesla already operates a small scale robotaxi service in Austin using supervised Model Ys, but the Cybercab is designed from the ground up for high-volume, low-cost production, with Musk stating an eventual goal of producing one vehicle every 10 seconds.

At Lucid Investor Day 2026, the company introduced Lunar, a purpose-built robotaxi concept based on the Midsize platform.
Into this landscape steps Lucid’s Lunar. Built on the company’s all-new Midsize EV platform, which will also underpin consumer SUVs starting below $50,000. The Lunar mirrors the Cybercab’s core philosophy of having two seats, no driver controls, and a focus on fleet economics. The platform introduces Lucid’s redesigned Atlas electric drive unit, engineered to be smaller, lighter, and cheaper to manufacture at scale.
Unlike Tesla’s strategy of building its own ride hailing network from scratch, Lucid is partnering with Uber. The companies are said to be in advanced discussions to deploy Midsize platform vehicles at large scale, with Uber CEO Dara Khosrowshahi publicly backing Lucid’s engineering credentials and autonomous-ready architecture.
In the investor day event, Lucid also outlined a recurring software revenue model, with an in-vehicle AI assistant and monthly autonomous driving subscriptions priced between $69 and $199. This can be seen as a nod to the software revenue stream that Tesla has long championed with its Full Self-Driving subscription.
Tesla’s Cybercab is targeting a price point below $30k and with operating costs as low as 20 cents per mile. But with regulatory hurdles still ahead, the window for competition is open. Lucid’s Lunar may not have a launch date yet, but it arrives at a pivotal moment, and when the robotaxi race is no longer viewed as hypothetical. Rather, every serious EV player needs to come to bat on the same plate that Tesla has had countless practice swings on over the last seven years.
Elon Musk
Brazil Supreme Court orders Elon Musk and X investigation closed
The decision was issued by Supreme Court Justice Alexandre de Moraes following a recommendation from Brazil’s Prosecutor-General Paulo Gonet.
Brazil’s Supreme Federal Court has ordered the closure of an investigation involving Elon Musk and social media platform X. The inquiry had been pending for about two years and examined whether the platform was used to coordinate attacks against members of the judiciary.
The decision was issued by Supreme Court Justice Alexandre de Moraes following a recommendation from Brazil’s Prosecutor-General Paulo Gonet.
According to a report from Agencia Brasil, the investigation conducted by the Federal Police did not find evidence that X deliberately attempted to attack the judiciary or circumvent court orders.
Prosecutor-General Paulo Gonet concluded that the irregularities identified during the probe did not indicate fraudulent intent.
Justice Moraes accepted the prosecutor’s recommendation and ruled that the investigation should be closed. Under the ruling, the case will remain closed unless new evidence emerges.
The inquiry stemmed from concerns that content on X may have enabled online attacks against Supreme Court justices or violated rulings requiring the suspension of certain accounts under investigation.
Justice Moraes had previously taken several enforcement actions related to the platform during the broader dispute involving social media regulation in Brazil.
These included ordering a nationwide block of the platform, freezing Starlink accounts, and imposing fines on X totaling about $5.2 million. Authorities also froze financial assets linked to X and SpaceX through Starlink to collect unpaid penalties and seized roughly $3.3 million from the companies’ accounts.
Moraes also imposed daily fines of up to R$5 million, about $920,000, for alleged evasion of the X ban and established penalties of R$50,000 per day for VPN users who attempted to bypass the restriction.
Brazil remains an important market for X, with roughly 17 million users, making it one of the platform’s larger user bases globally.
The country is also a major market for Starlink, SpaceX’s satellite internet service, which has surpassed one million subscribers in Brazil.
Elon Musk
FCC chair criticizes Amazon over opposition to SpaceX satellite plan
Carr made the remarks in a post on social media platform X.
U.S. Federal Communications Commission (FCC) Chairman Brendan Carr criticized Amazon after the company opposed SpaceX’s proposal to launch a large satellite constellation that could function as an orbital data center network.
Carr made the remarks in a post on social media platform X.
Amazon recently urged the FCC to reject SpaceX’s application to deploy a constellation of up to 1 million low Earth orbit satellites that could serve as artificial intelligence data centers in space.
The company described the proposal as a “lofty ambition rather than a real plan,” arguing that SpaceX had not provided sufficient details about how the system would operate.
Carr responded by pointing to Amazon’s own satellite deployment progress.
“Amazon should focus on the fact that it will fall roughly 1,000 satellites short of meeting its upcoming deployment milestone, rather than spending their time and resources filing petitions against companies that are putting thousands of satellites in orbit,” Carr wrote on X.
Amazon has declined to comment on the statement.
Amazon has been working to deploy its Project Kuiper satellite network, which is intended to compete with SpaceX’s Starlink service. The company has invested more than $10 billion in the program and has launched more than 200 satellites since April of last year.
Amazon has also asked the FCC for a 24-month extension, until July 2028, to meet a requirement to deploy roughly 1,600 satellites by July 2026, as noted in a CNBC report.
SpaceX’s Starlink network currently has nearly 10,000 satellites in orbit and serves roughly 10 million customers. The FCC has also authorized SpaceX to deploy 7,500 additional satellites as the company continues expanding its global satellite internet network.