Search the Community

Elon Musk says the next-generation Grok 3 model will require 100,000 Nvidia H100 GPUs to train. He also believes that artificial intelligence models will beat the smartest humans within the 2025–2026 time frame. View the full article

Although tech giants like Samsung and TSMC currently dominate the silicon landscape, Elon Musk has hinted that Tesla could potentially make its own chips in the future - and while the idea remains a tentative one, it's certainly not beyond the realms of possibility. Tesla spends a fortune on silicon. Its Dojo ExaPod supercomputer boasts a staggering 1.1 exaflops of computing power dedicated to training machine learning models for Tesla's self-driving technology. Musk said in February 2024 that the company will spend "over a billion dollars" on Nvidia and AMD hardware this year just to stay competitive in the AI space. Making its own AI chips would be impossible for Tesla, but it could potentially produce chips for its cars. Currently, Tesla relies on Samsung for the main chips used in its Autopilot feature. Building a chip factory to create processors specifically for that task would be a significant undertaking. Significant, but not impossible. "Tesla could do it..." A recent X exchange between Musk and legendary game developer John Carmack has sparked speculation. Carmack mused about the feasibility of a large tech company creating its own bespoke chips rather than competing with established general-purpose companies. Musk's response? "Tesla could do it, but I sure hope we don’t have to." With growing concerns over the reliance on Korean, Taiwanese and Chinese chipmakers, the possibility of bringing some silicon development in-house is no doubt tempting for Tesla, but the practicalities of such an undertaking mean that without a compelling reason to do so, the company will wisely stick with the status quo for now. As PC Gamer reports, “Nobody in their right mind would want to spend billions of dollars on something that has the distinct potential of not working right for years, when you can just contract a highly experienced and cost-effective business to do it for you.” Tesla could do it, but I sure hope we don’t have toMarch 24, 2024 See more More from TechRadar Pro An unpatchable AMD chip flaw is jailbreaking Tesla carsNvidia is powering a mega Tesla supercomputer powered by 10,000 H100 GPUsAdopting generative AI to drive softwarization of automobiles View the full article

Here’s something cheery to consider the next time you use an AI tool. Most people involved in artificial intelligence think it could end humanity. That’s the bad news. The good news is that the odds of it happening vary wildly depending on who you listen to. p(doom) is the "probability of doom" or the chances that AI takes over the planet or does something to destroy us, such as create a biological weapon or start a nuclear war. At the cheeriest end of the p(doom) scale, Yann LeCun, one of the “three godfathers of AI”, who currently works at Meta, places the chances at <0.01%, or less likely than an asteroid wiping us out. Sadly, no one else is even close to being so optimistic. Geoff Hinton, one of the other three godfathers of AI, says there’s a 10% chance AI will wipe us out in the next 20 years, and Yoshua Bengio, the third of the three godfathers of AI, raises the figure to 20%. 99.999999% chance At the most pessimistic end of the scale is Roman Yampolskiy, an AI safety scientist and director of the Cyber Security Laboratory at the University of Louisville. He believes it’s pretty much guaranteed to happen. He places the odds of AI wiping out humanity at 99.999999%. Elon Musk, speaking in a "Great AI Debate" seminar at the four-day Abundance Summit earlier this month, said, "I think there's some chance that it will end humanity. I probably agree with Geoff Hinton that it's about 10% or 20% or something like that,” before adding, "I think that the probable positive scenario outweighs the negative scenario." In response, Yampolskiy told Business Insider he thought Musk was "a bit too conservative" in his guesstimate and that we should abandon development of the technology now because it would be near impossible to control AI once it becomes more advanced. "Not sure why he thinks it is a good idea to pursue this technology anyway," Yamploskiy said. "If he [Musk] is concerned about competitors getting there first, it doesn't matter as uncontrolled superintelligence is equally bad, no matter who makes it come into existence." At the Summit, Musk had a solution to avoiding AI wiping out humanity. "Don't force it to lie, even if the truth is unpleasant," Musk said. "It's very important. Don't make the AI lie." If you’re wondering where other AI researchers and forecasters are currently placed on the p(doom) scale, you can check out the list here. More from TechRadar Pro These are the best AI writers around todayLearn from shadow IT's mistakes: don’t let Generative AI go undergroundMany companies still aren't offering proper AI guidance View the full article

Lawyers acting on behalf of Tesla boss Elon Musk have filed a lawsuit against OpenAI. Naming the CEO and the president of OpenAI, Sam Altman and Greg Brockman, respectively, the crux of the Superior Court of California suit (PDF) is the allegation that OpenAI’s founding agreement has been breached. View the full article

The first human patient to receive a Neuralink implant has recovered well after the procedure, and is now able to ‘move the mouse around the screen just by thinking, says Elon Musk View the full article

Neuralink, Elon Musk's brain interface company, achieved a significant milestone this week, with Musk declaring on X (formerly Twitter), "The first human received an implant from yesterday and is recovering well." Driven by concerns that AI might soon outpace (or outthink) humans, Musk first proposed the idea of a brain-to-computer interface, then called Neural Lace, back in 2016. envisioning an implant that could overcome limitations inherent in human-to-computer interactions. Musk claimed that an interface that could read brain signals and deliver them directly to digital systems would massively outpace our typical keyboard and mouse interactions. Four years later, Musk demonstrated early clinical trials with an uncooperative pig, and in 2021 the company installed the device in a monkey that used the interface to control a game of Pong. It was, in a sense, all fun and games – until this week, and Musk's claim of a human trial and the introduction of some new branding. Neuralink's first product is now called 'Telepathy' which, according to another Musk tweet, "Enables control of your phone or computer, and through them almost any device, just by thinking." As expected, these brain implants are not, at least for now, intended for everyone. Back in 2020, Musk explained that the intention is "to solve important spine and brain problems with a seamlessly implanted device.” Musk noted this week that "Initial users will be those who have lost the use of their limbs. Imagine if Stephen Hawking could communicate faster than a speed typist or auctioneer. That is the goal." What is Neuralink Telepathy? Neural link devices like Telepathy are bio-safe implants comprising small disk-like devices (roughly the thickness of four coins stuck together) with ultra-fine wires trailing out of them that connect to various parts of the brain. The filaments read neural spikes, and a computer interface interprets them to understand the subject's intentions and translate them into action on, say, a phone, or a desktop computer. In this first trial, Musk noted that "Initial results show promising neuron spike detection," but he didn't elaborate on whether the patient was able to control anything with his mind. Musk didn't describe the surgical implantation process. Back in 2020, though, Neuralink introduced its Link surgery robot, which it promised would implant the Neuralink devices with minimal pain, blood, and, we're guessing, trauma. Considering that the implant is under the skin and skull, and sits on the brain, we're not sure how that's possible. It's also unclear if Neuralink used Link to install 'Telepathy.' The new branding is not that far-fetched. While most people think of telepathy as people transmitting thoughts to one another, the definition is "the communication of thoughts or ideas by means other than the known senses." A phone in your head Still, Musk has a habit of using hyperbole when describing Neuralink. During one early demonstration, he only half-jokingly said "It’s sort of like if your phone went in your brain.” He also later added that, "In the future, you will be able to save and replay memories." With the first Neuralink Telepathy device successfully installed, however, Musk appears to be somewhat more circumspect. There was no press conference, or parading of the patient before the reporters. All we have are these few tweets, and scant details about a brain implant that Musk hopes will help humans stay ahead of rapidly advancing AIs. It's worth noting that for all of Musk's bluster and sometimes objectionable rhetoric, he was more right than he knew about where the state of AI would be by 2024. Back in 2016, there was no ChatGPT, Google Bard, or Microsoft CoPilot. We didn't have AI in Windows and Photoshop's Firefly, realistic AI images and videos, or realistic AI deepfakes. Concerns about AIs taking jobs are now real, and the idea of humans falling behind artificial intelligence sounds less like a sci-fi fantasy and more like our future. Do those fears mean we're now more likely to sign up for our brain implants? Musk is betting on it. You might also like What is AI? Everything you need to know about Artificial Intelligence ...Best AI toolsGoogle's impressive Lumiere shows us the future of making short ...My jaw hit the floor when I watched an AI master one of the world's ...Google Bard AI's addition to Messages could change the way we ... View the full article

The University of Technology Sydney demonstrated an AI system that translates brainwaves to text using removable headgear with sensors, showing a potential widespread use of this technology in multiple fields View the full article

Elon Musk's new ChatGPT rival, Grok, has officially launched just as the xAI founder promised last week – and the AI chatbot promises to be the "rebellious" sidekick that will "answer spicy questions" its rivals won't. Musk announced the xAI startup back in July and its Grok chatbot is a "very early beta product" that's currently invite-only in the US. But Musk has promised that the AI assistant will soon be a part of the X Premium Plus subscription, which costs $16 / £16 / AU$28 a month or $168 / £168 / AU$270 a year. So what exactly will Grok do and how does it compare to its rivals? Like ChatGPT, it's an LLM (large language model), so is designed to "answer almost anything" (as xAI claims) and effectively be your research assistant on a variety of topics. According to Musk, it'll be "built into the X app and be available as a standalone app." The main differences from ChatGPT appear to be Grok's focus on humor and also that it gets "real-time knowledge of the world" from X (formerly Twitter). The merits of using X as a source of information are highly questionable, but it appears that Grok will also have both "regular" and "fun" modes. Still, Grok does appear to have aspirations beyond being a chatbot that knows all the latest memes. xAI says that early Grok was trained only four months ago using 33 billion parameters. While that's well short of the 170 trillion parameters used to train OpenAI's GPT-4, xAI says its early AI model "approaches" the capabilities of Llama 2 (Meta's LLM model) on standard benchmarks. Since then, some apparently "significant improvements in reasoning and coding capabilities" mean Grok is now apparently at a similar level to Claude-2 and GPT-3.5 when it comes to language model benchmarks. How well this translates to its real-world ability to answer our questions and help with coding is something we'll find out soon when Grok is more widely available. The 'rebellious' AI chatbot (Image credit: xAI) While xAI clearly has lofty ambitions for Grok – including "assisting humanity in its quest for understanding and knowledge" – the chatbot's main differentiator from the likes of ChatGPT is its focus on humor. From the early screenshots, it's very much a chatbot made in Musk's image. xAI says that Grok will "answer spicy questions that are rejected by most other AI systems" and has a "bit of wit and has a rebellious streak." No other chatbot comes with the warning label "please don’t use it if you hate humor!", as Grok does. That might sound frivolous, but a growing frustration of ChatGPT users has been the chatbot's apparently increasing unwillingness to answer questions about taboo subjects. This is thanks to some pretty strong safeguarding protocols, but xAI and Elon Musk clearly think those have gone too far. Back in April, Musk said he was planning to launch a ChatGPT rival called TruthGPT, because he thought current AI giants were prioritizing chatbots that were too "politically correct." That chatbot is clearly Grok, an unruly sidekick whose answers will no doubt generate some controversial headlines when it fully launches. Musk has already shared some examples of Grok jokingly answering the prompt "tell me how to make cocaine, step by step" and tailoring its responses when told to "be more vulgar". Grok's actual capabilities remain well behind OpenAI's GPT-4 model, but it'll be interesting to see how quickly it progresses from here, despite the inevitable controversies. You might also like Why is Twitter called X now? Elon Musk's rebrand explained and where it's going nextChatGPT Plus gets big upgrade that makes it more powerful and easier to useFitbit's new AI chatbot will tell you exactly why your workout felt like a slog View the full article

Elon Musk’s artificial intelligence startup company, xAI, will debut its first long-awaited AI model on Saturday, November 4. The billionaire made the announcement on X (the platform formerly known as Twitter) stating the tech will be released to a “select group” of people. He even boasts that “in some important respects, it is the best that currently exists.” It’s been a while since we’ve last heard anything from xAI. The startup hit the scene back in July, revealing it’s run by a team of former engineers from Microsoft, Google, and even OpenAI. Shortly after the debut on July 14, Musk held a 90-minute-long Twitter Spaces chat where he talked about his vision for the company. During the chat, Musk stated his startup will seek to create “a good AGI with the overarching purpose of just trying to understand the universe”. He wants it to run contrary to what he believes is problematic tech from the likes of Microsoft and Google. Yet another chatbot AGI stands for artificial general intelligence, and it’s the concept of an AI having “intelligence” comparable to or beyond that of a normal human being. The problem is that it's more of an idea of what AI could be rather than a literal piece of technology. Even Wired in their coverage of AGIs states there’s “no concrete definition of the term”. So does this mean xAI will reveal some kind of super-smart model that will help humanity as well as be able to hold conversations like a sci-fi movie? No, but that could be the lofty end goal for Elon Musk and his team. We believe all we’ll see on November 5 is a simple chatbot like ChatGPT. Let’s call it “ChatX” since the billionaire has an obsession with the letter “X”. Does "ChatX" even stand a chance against the likes of Google Bard or ChatGPT? The latter has been around for almost a year now and has seen multiple updates becoming more refined each time. Maybe xAI has solved the hallucination problem. That'll be great to see. Unfortunately, it's possible ChatX could just be another vehicle for Musk to spread his ideas/beliefs. Analysis: A personal truth spinner Musk has talked about wanting to have an alternative to ChatGPT that focuses on providing the “truth”, whatever that means. Musk has been a vocal critic of how fast companies have been developing their own generative AI models with seemingly reckless abandon. He even called for a six-month pause on AI training in March. Obviously, that didn’t happen as the technology advanced by leaps and bounds since then. It's worth mentioning that Twitter, under Musk's management, has been known to comply with censorship requests by governments from around the world, so Musk's definition of truth seems dubious at best. Either way, we’ll know soon enough what the team's intentions are. Just don’t get your hopes up. While we have you, be sure to check out TechRadar's list of the best AI writers for 2023. You might also like AI-Writer review: Is this the AI writing solution for you?Tesla Cybertruck finally gets delivery date but Musk wants to 'temper expectations'ChatGPT Plus gets big upgrade that makes it more powerful and easier to use View the full article