Tesla CEO Elon Musk said on Friday that he plans for his newest venture, the artificial intelligence startup xAI, to collaborate with the automaker both on the “silicon front” and on the “AI software front.”
Musk also said, during Friday’s live audio session on Twitter Spaces, that xAI will use Twitter data for training the “maximally curious” artificial intelligence systems and products he hopes to build there. Musk did not specify whether and how much Twitter will charge xAI or his other companies for its data.
related investing news
When Musk led a buyout of the social media venture in October 2022, Twitter took on $13 billion in new debt. The company has struggled to juice its subscription revenue, and has been sued by ex-employees and vendors for non-payment for completed work or severance.
Several of the other companies where Musk was a founder or serves as CEO, including Tesla, SpaceX and The Boring Co., have done business together for years. Some of their transactions have been disclosed in Tesla financial filings with the U.S. Securities and Exchange Commission.
On Friday, without citing evidence, Musk alleged that “Every AI organization on Earth” had used Twitter’s data for training, “in all cases illegally.” It was not clear which laws would have been violated by others’ data scraping. Earlier this month, Twitter sued four unknown parties for data scraping in Texas.
Twitter implemented rate limits on the social media platform in recent weeks because, Musk claimed, it was “being scraped like crazy.” He said, “We had multiple entities scraping every tweet ever made, and trying to do so in like, basically a span of days. So — this was bringing the system to its knees. So we had to take action.” He apologized for the inconvenience of the rate limiting.
In light of widespread use of Twitter data by AI software developers, Musk said, “I guess we will use the public tweets — obviously not anything private — for training as well, just like basically everyone else has.”
Twitter’s data set appeals for “text training,” and “image and video training,” Musk said. However, he specified that AI systems need more than human-created data and he was hoping that xAI could follow in the footsteps of Alphabet-owned DeepMind’s Alpha Zero, a computer program that achieved a masterful level of play in three games, chess, shogi and go, after training by playing these games against itself.
A Tesla fan and promoter, Omar Qazi (known as Whole Mars Catalog on Twitter) asked Musk a few questions about how he plans for xAI to work with Tesla during the Spaces event. Among other things, he asked whether xAI would potentially use Nvidia- or Tesla-made silicon for data processing.
Musk said, “That’s sort of a Tesla question. Tesla is building custom silicon. I wouldn’t call anything that Tesla’s producing a ‘GPU’ although one can characterize it in GPU equivalents.” He then spoke about Tesla’s in-vehicle hardware, which enables the company’s advanced driver assistance systems to work in its cars. The systems are marketed as Autopilot and Full Self Driving capability in the US.
Tesla has been promising fans a robotaxi, or self-driving vehicle, for years. At that time, Musk said a cross-country demo with a Tesla car would be possible without a single human intervention by the end of 2017. In 2019, Tesla raised billions of dollars with the promise of a million robotaxi-ready Tesla vehicles on the road in a year. So far, none of Tesla’s vehicles are capable of operating without a human driver ready to steer or brake at any time.
Musk said on Twitter Spaces on Friday that Tesla’s hardware 4, which is shipping in now, is “three-to-five times more capable than hardware 3,” and promised “hardware 5” would come along in a few years and would be “four or five times more capable” than its current version.
The CEO also discussed Dojo, a supercomputer Tesla is developing for AI machine learning and computer vision training purposes. Tesla uses video clips and data from its customers’ vehicles to improve existing software, or develop new features.
Musk said that the eventual AI language model that xAI will presumably develop won’t be “politically correct.” The CEO, who has repeatedly attacked “woke” or progressive values, said “I think our AI can give answers that people may find controversial even though they are actually true.”
The Tesla CEO said that xAI will need to develop technology that “understands the physical world and not just the Internet,” and he thinks that Tesla’s driving data will help it on that front.
Walter Isaacson, the author of an Elon Musk biography coming out later this year, asked Musk about Optimus, a humanoid robot Tesla is developing with the aim of using it in manufacturing. Musk said that the robot is still in its “early stages” and his team needs to find a way that users will be able to easily turn it off.
Tesla showed off a design for a humanoid robot called Optimus at its AI day in September 2022. Tesla executive are expected to share updates on this and more on an earnings call next Wednesday.
Elon Musk’s health tech company Neuralink labeled itself a “small disadvantaged business” in a federal filing with the U.S. Small Business Administration, shortly before a financing round valued the company at $9 billion.
Neuralink is developing a brain-computer interface (BCI) system, with an initial aim to help people with severe paralysis regain some independence. BCI technology broadly can translate a person’s brain signals into commands that allow them to manipulate external technologies just by thinking.
Neuralink’s filing, dated April 24, would have reached the SBA at a time when Musk was leading the Trump administration’s Department of Government Efficiency. At DOGE, Musk worked to slash the size of federal agencies.
MuskWatch first reported on the details Neuralink’s April filing.
According to the SBA’s website, a designation of SDB means a company is at least 51% owned and controlled by one or more “disadvantaged” persons who must be “socially disadvantaged and economically disadvantaged.” An SDB designation can also help a business “gain preferential access to federal procurement opportunities,” the SBA website says.
Musk, the world’s wealthiest person, is CEO of Tesla and SpaceX, in addition to his other businesses like artificial intelligence startup xAI and tunneling venture The Boring Company. In 2022, Musk led the $44 billion purchase of Twitter, which he later named X before merging it with xAI.
Jared Birchall, a Neuralink executive, was listed as the contact person on the filing from April. Birchall, who also manages Musk’s money as head of his family office, didn’t immediately respond to a request for comment.
Neuralink, which incorporated in Nevada, closed a $650 million funding round in early June at a $9 billion valuation. ARK Invest, Peter Thiel’s Founders Fund, Sequoia Capital and Thrive Capital were among the investors. Neuralink said the fresh capital would help the company bring its technology to more patients and develop new devices that “deepen the connection between biological and artificial intelligence.”
Under Musk’s leadership at DOGE, the initiative took aim at government agencies that emphasized diversity, equity and inclusion (DEI). In February, for example, DOGE and Musk boasted of nixing hundreds of millions of dollars worth of funding for the Department of Education that would have gone towards DEI-related training grants.
Defense manufacturing startup Hadrian on Thursday announced the closing of $260 million Series C funding round led by Peter Thiel‘s Founders Fund and Lux Capital.
The machine parts company said it will use the funding to build a new 270,000 square foot factory in Mesa, Arizona, and expand its Torrance, California, location as it looks to beef up its shipbuilding and naval defense capabilities.
“What we really need in this country is this quantum leap above China’s manufacturing model,” said CEO Chris Power in an interview with CNBC’s Morgan Brennan. “It’s about supercharging the worker versus replacing them.”
Defense tech startups like Hadrian are disrupting the mainstay defense contracting industry, which is led by leaders such as Northrop Grumman and Lockheed Martin, and battling it out to boost U.S. defense production while scooping up Department of Defense contracts.
An overall view of the manufacturing line in a Hadrian Automation Inc. factory.
Courtesy: Hadrian Automation, Inc.
Hadrian said the Arizona space will be four times the size of its California facility and start operations by Christmas. The factory will create 350 local jobs. The Hawthrone, California-based company said it is working on four to five new facilities to support production over the next year to support Department of Defense needs.
Read more CNBC tech news
Hadrian said it uses robotics and artificial intelligence to automate factories that can “supercharge American workers.”
Power said demand is rapidly growing, but the lack of U.S.-based talent is a major hurdle to building American dominance in shipbuilding and submarines.
Using its tools, the company said it can train workers within 30 days, making them 10 times more productive. Its workforce includes ex-marines and former nurses who have never set foot in a factory.
An overall view of the manufacturing line in a Hadrian Automation Inc. factory.
Courtesy: Hadrian Automation, Inc.
“We have to do a lot more … but certainly we’re able to keep up with the scale right now, and grateful to our team and customers for letting us go and do that,” he said. “As a country, we have to treat this like a national security crisis, not just the economics of manufacturing.”
The fresh raise also includes investments from Andreessen Horowitz and new stakeholders such as Brad Gerstner’s Altimeter Capital.
The company closed a $92 million funding round in late 2023.
Attendees walk through an exposition hall at AWS re:Invent, a conference hosted by Amazon Web Services, in Las Vegas on Dec. 3, 2024.
Noah Berger | Getty Images
Amazon is laying off some staffers in its cloud computing division, the company confirmed on Thursday.
“After a thorough review of our organization, our priorities, and what we need to focus on going forward, we’ve made the difficult business decision to eliminate some roles across particular teams in AWS,” Amazon spokesperson Brad Glasser said in a statement. “We didn’t make these decisions lightly, and we’re committed to supporting the employees throughout their transition.”
The company declined to say which units within Amazon Web Services were impacted, or how many employees will be let go as a result of the job cuts.
Reuters was first to report on the layoffs.
In May, Amazon reported a third straight quarterly revenue miss at AWS. Sales increased 17% to $29.27 billion in the first quarter, slowing from 18.9% in the prior period.
Amazon said the cuts weren’t primarily due to investments in artificial intelligence, but are a result of ongoing efforts to streamline the workforce and refocus on certain priorities. The company said it continues to hire within AWS.
Amazon CEO Andy Jassy has been on a cost-cutting mission for the past several years, which has resulted in more than 27,000 employees being let go since 2022. Job reductions have continued this year, though at a smaller scale than preceding years. Amazon’s stores, communications and devices and services divisions have been hit with layoffs in recent months.
AWS last year cut hundreds of jobs in its physical stores technology and sales and marketing units.
Last month, Jassy predicted that Amazon’s corporate workforce could shrink even further as a result of the company embracing generative AI.
“We will need fewer people doing some of the jobs that are being done today, and more people doing other types of jobs,” Jassy told staffers. “It’s hard to know exactly where this nets out over time, but in the next few years, we expect that this will reduce our total corporate workforce.”