Elon Musk has debunked the claims of a recent Wall Street Journal report which suggested that Tesla and xAI have discussed a potential deal where the electric vehicle maker would license the AI startup’s models to help develop technologies such as Full Self-Driving (FSD). As per Musk, there is no need for Tesla to license anything from xAI.
Citing people reportedly familiar with the matter, the WSJ noted that the proposed deal between Tesla and xAI has been described to investors. With the deal in place, Tesla would reportedly license xAI’s artificial intelligence models to help power FSD. Tesla would then share some of FSD’s revenue with xAI.
The Journal’s report also alleged that xAI would be developing other features for Tesla’s vehicles, such as a Siri-esque voice assistant and the software that would power Optimus, the company’s humanoid robot.
NEWS: Elon Musk’s AI startup @xAI has discussed a deal where it would get some @Tesla revenue in exchange for providing the carmaker access to its technology and resources, the latest example of the growing interconnectedness of Musk’s companies.
Under a proposed arrangement as… pic.twitter.com/SBmlOT3qRU— ALEX (@ajtourville) September 8, 2024
In a post on X, Musk noted that while he has not fully read the WSJ report and all its points, the article’s overall claim is inaccurate. Musk admitted that Tesla has learned a lot from discussions with xAI’s engineers, but there is really no need to license anything from the AI startup.
Musk also explained that there are massive differences between the AI models that are being developed by xAI and Tesla. xAI’s models are gigantic and could not possibly run on Tesla’s electric vehicles, while the EV maker’s models have incredibly dense intelligence since they are focused on real-world driving.
“Haven’t read the article, but the above is not accurate. Tesla has learned a lot from discussions with engineers at xAI that have helped accelerate achieving unsupervised FSD, but there is no need to license anything from xAI.
Haven’t read the article, but the above is not accurate.
Tesla has learned a lot from discussions with engineers at xAI that have helped accelerate achieving unsupervised FSD, but there is no need to license anything from xAI.
The xAI models are gigantic, containing, in…— Elon Musk (@elonmusk) September 8, 2024
“The xAI models are gigantic, containing, in compressed form, most of human knowledge, and couldn’t possibly run on the Tesla vehicle inference computer, nor would we want them to. The Tesla AI models have incredibly “dense” (in a good way, lol) intelligence, as they compress video of reality into driving commands, but must operate on a ~300W computer with memory size and bandwidth far lower than, say, an H100 GPU.
“Tesla real-world AI also has a vastly larger context size than an LLM, as the combined video history from all cameras is several gigabytes in size,” Musk wrote.
Don’t hesitate to contact us with news tips. Just send a message to simon@teslarati.com to give us a heads up.
Tesla is being accused of infringing robotics patents by a company called Perrone Robotics, which is based out of Charlottesville, Virginia.
The suit was filed in Alexandria, Virginia, and accuses Tesla of knowingly infringing upon five patents related to robotics systems for self-driving vehicles.
The company said its founder, Paul Perrone, developed general-purpose robotics operating systems for individual robots and automated devices.
Perrone Robotics claims that all Tesla vehicles utilizing the company’s Autopilot suite within the last six years infringe the five patents, according to a report from Reuters.
Tesla’s new Safety Report shows Autopilot is nine times safer than humans
One patent was something the company attempted to sell to Tesla back in 2017. The five patents cover a “General Purpose Operating System for Robotics,” otherwise known as GPROS.
The GPROS suite includes extensions for autonomous vehicle controls, path planning, and sensor fusion. One key patent, U.S. 10,331,136, was explicitly offered to Tesla by Perrone back in 2017, but the company rejected it.
The suit aims to halt any further infringements and seeks unspecified damages.
This is far from the first suit Tesla has been involved in, including one from his year with Perceptive Automata LLC, which accused Tesla of infringing on AI models to interpret pedestrian/cyclist intent via cameras without licensing. Tesla appeared in court in August, but its motion to dismiss was partially denied earlier this month.
Tesla also settled a suit with Arsus LLC, which accused Autopilot’s electronic stability features of infringing on rollover prevention tech. Tesla won via an inter partes review in September.
Most of these cases involve non-practicing entities or startups asserting broad autonomous vehicle patents against Tesla’s rapid iteration.
Tesla typically counters with those inter partes reviews, claiming invalidity. Tesla has successfully defended about 70 percent of the autonomous vehicle lawsuits it has been involved in since 2020, but settlements are common to avoid discovery costs.
The case is Perrone Robotics Inc v Tesla Inc, U.S. District Court, Eastern District of Virginia, No. 25-02156. Tesla has not yet listed an attorney for the case, according to the report.
News
Tesla has passed a critical self-driving milestone Elon Musk listed in Master Plan Part Deux
Tesla China announced that the company’s Autopilot system has accumulated 10 billion kilometers of driving experience.
Tesla has passed a key milestone, and it was one that CEO Elon Musk initially mentioned more than nine years ago when he published Master Plan, Part Deux.
As per Tesla China in a post on its official Weibo account, the company’s Autopilot system has accumulated over 10 billion kilometers of real-world driving experience.
Tesla China’s subtle, but huge announcement
In its Weibo post, Tesla China announced that the company’s Autopilot system has accumulated 10 billion kilometers of driving experience. “In this respect, Tesla vehicles equipped with Autopilot technology can be considered to have the world’s most experienced and seasoned driver.”
Tesla AI’s handle on Weibo also highlighted a key advantage of the company’s self-driving system. “It will never drive under the influence of alcohol, be distracted, or be fatigued,” the team wrote. “We believe that advancements in Autopilot technology will save more lives.”
Tesla China did not clarify exactly what it meant by “Autopilot” in its Weibo post, though the company’s intense focus on FSD over the past years suggests that the term includes miles that were driven by FSD (Beta) and Full Self-Driving (Supervised). Either way, 10 billion cumulative miles of real-world data is something that few, if any, competitors could compete with.
Elon Musk’s 10-billion-km estimate, way back in 2016
When Elon Musk published Master Plan Part Deux, he outlined his vision for the company’s autonomous driving system. At the time, Autopilot was still very new, though Musk was already envisioning how the system could get regulatory approval worldwide. He estimated that worldwide regulatory approval will probably require around 10 billion miles of real-world driving data, which was an impossible-sounding amount at the time.
“Even once the software is highly refined and far better than the average human driver, there will still be a significant time gap, varying widely by jurisdiction, before true self-driving is approved by regulators. We expect that worldwide regulatory approval will require something on the order of 6 billion miles (10 billion km). Current fleet learning is happening at just over 3 million miles (5 million km) per day,” Musk wrote.
It’s quite interesting but Tesla is indeed getting regulatory approval for FSD (Supervised) at a steady pace today, at a time when 10 billion miles of data has been achieved. The system has been active in the United States and has since been rolled out to other countries such as Australia, New Zealand, China, and, more recently, South Korea. Expectations are high that Tesla could secure FSD approval in Europe sometime next year as well.
Elon Musk
SpaceX maintains unbelievable Starship target despite Booster 18 incident
It appears that it will take more than an anomaly to stop SpaceX’s march towards Starship V3’s refinement.
SpaceX recently shared an incredibly ambitious and bold update about Starship V3’s 12th test flight.
Despite the anomaly that damaged Booster 18, SpaceX maintained that it was still following its plans for the upgraded spacecraft and booster for the coming months. Needless to say, it appears that it will take more than an anomaly to stop SpaceX’s march towards Starship V3’s refinement.
Starship V3 is still on a rapid development path
SpaceX’s update was posted through the private space company’s official account on social media platform X. As per the company, “the Starbase team plans to have the next Super Heavy booster stacked in December, which puts it on pace with the test schedule planned for the first Starship V3 vehicle and associated ground systems.”
SpaceX then announced that Starship V3’s maiden flight is still expected to happen early next year. “Starship’s twelfth flight test remains targeted for the first quarter of 2026,” the company wrote in its post on X.
Elon Musk mentioned a similar timeline on X earlier this year. In the lead up to Starshp Flight 11, which proved flawless, Musk stated that “Starship V3 is a massive upgrade from the current V2 and should be through production and testing by end of year, with heavy flight activity next year.” Musk has also mentioned that Starship V3 should be good enough to use for initial Mars missions.
Booster 18 failure not slowing Starship V3’s schedule
SpaceX’s bold update came after Booster 18 experienced a major anomaly during gas system pressure testing at SpaceX’s Massey facility in Starbase, Texas. SpaceX confirmed in a post on X that no propellant was loaded, no engines were installed, and personnel were positioned at a safe distance when the booster’s lower section crumpled, resulting in no injuries.
Still, livestream footage showed significant damage around the liquid oxygen tank area of Booster 18, leading observers to speculate that the booster was a total loss. Booster 18 was among the earliest vehicles in the Starship V3 series, making the failure notable. Despite the setback, Starship V3’s development plans appear unchanged, with SpaceX pushing ahead of its Q1 2026 test flight target.
Tesla accused of infringing robotics patents in new lawsuit
Tesla has passed a critical self-driving milestone Elon Musk listed in Master Plan Part Deux
