Elon Musk’s TESLA AI Just Crushed OpenAI – Is This the End of ChatGPT?

Rock 3. Let me tell you, it's insane. It's super fast, really smart and scary good at getting things right. I bet Chad GBT is sweating right now. Love him or hate him, Elon Musk sure knows how to grab attention when he launches something. And now he's teaming up with Donald Trump for something massive. We go by all the news announcements. Looks like the Tesla AI we've been waiting for. So, that raises the question, what will a Tesla AI actually look like? Today we're checking out Elon's background in AI, his connection to Chad GBT, what he's working on right now, and where he's going next. This is going to be wild. Elon's journey with AI really

got going in 2014 when he became one of the first backers of DeepMind, an AI research company that Google ended up buying that same year. Then in December 2015, he co-founded OpenAI. The goal was to set it up as a nonprofit focused on building AGI, artificial general intelligence. But even from day one, there was drama, mostly about money. OpenAI needed funding, and yes, they did manage to get some. Elon put in less than $45 million, and other investors brought in over $90 million. When OpenAI was just starting, Greg Brockman and Sam Alman planned to raise $100 million total. Then Elon wrote an email saying, "We need a much bigger number than $100 million so we don't look hopeless. I think we should say we have $1 billion funding commitment. " That came straight from an internal email OpenAI released last year. Even though they took care of

the money issue, the vision wasn't totally in sync. Elon's aim was to make sure AI advancements would help humanity and stand up to big tech AI project. Meanwhile, Sam put a bigger focus on openness and safety. These might seem like two sides of the same coin, but they actually point to different beliefs. Elon wanted to see the tech serve people and improve lives while still given some capitalist ideas like snagging part of the market. Sam and Greg, on the other hand, were creating the technology more out of pure passion, believing that AI should ultimately belong to everyone. As you can imagine

running a company without a clear direction was tough and everyone knew OpenAI needed to go forprofit to keep enough money coming in. But when they talked about switching to forprofit model, there was a big disagreement. OpenAI leaders felt that given any single person total control clashed with their mission. Elon on the other hand wanted majority ownership, big say on the board and the CEO title. He even suggested folding OpenAI into Tesla. In early February 2018, he forwarded an email saying OpenAI should attach to Tesla as its cash cow, adding that Tesla was the only path that could even hope to challenge Google, though the odds were still small. Elon officially stepped down from OpenAI's board in February 2018, saying it conflicted with Tesla's own AI plans. Let's be real, Tesla was putting a huge effort into self-driving cars and had even hired OpenAI's top researcher, Andre Karpathy, to lead autopilot. So, it's not shocking that Elon thought, "Okay, maybe it's time to walk away. " But listen, Elon Musk was a huge hit to OpenAI. They didn't just lose their most famous board

member, he also pulled a large donation he had promised. Even after stepping away, Elon couldn't resist throwing jabs at OpenAI. In 2020, he said they should be more open and admitted he didn't trust their leadership much when it came to AI safety, adding he wasn't very confident in Daario for that. Then in November 2022, after Chad GBT launched and blew up overnight, Elon got even louder. Just days after it went live, he accused OpenAI of misusing Twitter data. Remember, he had just bought Twitter around that time. This non-stop drive to keep pushing forward is part of why people are drawn to Elon. always growing, improving, learning. I bet he wishes he had access to Geek Academy back in 2022 with all its educational PDFs, toll collections, exclusive deals. Yeah, by early 2023, Elon was basically saying OpenAI had abandoned its original mission. In February 2023, once Microsoft got tighter with OpenAI, he reminded everyone OpenAI was created as an open-source. That's why it's called Open AI, nonprofit company to serve as a counterweight to Google, but now has become a closed source maximum profit company, effectively controlled by Microsoft. Not what I intended at all. Ouch. Then in March 2023, he and over a thousand other well-known figures signed a letter calling for a six-month pause on training AI beyond GBT4, labeling it a profound risk to society. One month later, he announced plans for truth GBT, self-described maximum truth seeking AI that tries to understand the nature of the universe, his direct answer to Chad GBT, which he slammed for being too politically correct. That truth GPT idea

eventually led to Elon's new AI company, XAI, already worth billions. This is where Elon's big vision and go big or go home style really shine. For starters, XAI managed to build and release Grock version 1 with just 12 team members. That alone shows they're serious. On top of that, Elon wants his AI to be less biased and more grounded in reality, maximally curious and truth seeking. His goal is for X AI to create what he calls good AGI, one that can stand up to major players like Google's deep mind and of course open AI. He believes those top AI labs aren't fully thinking about the risks of advanced AI. If you know Elon, you know this all ties back to his huge faith in humanity and his burning desire to get us to Mars someday. In a tech world full of super smart people and giant companies chasing big profits, Elon's style really does stand out. You might love him, you might not, but there is no denying that he's shaking things up. Will be fascinating to see how XAI fits into AI's next big chapter. But all

that was just talk and even more talk. What has Elon done practically? Something we all can touch and see. Well, Grock would be the obvious answer. When Grock one first launched, it caused a real buzz. Even though it was early days, XAI hyped up its personality, calling it a chatbot with a sense of humor and realtime data access through X. That was a not so subtle jab at Chad GBT. Elon basically hinted that Grock would deliver fresher, less sanitized responses. But the truly mind-blowing part was what someone at XAI said about Grock 1. It was the best we could do with 2 months of training and would improve rapidly with each passing week. Can you believe that? In just a couple of weeks, they had something that actually worked. Then came Grog 2, which only built on that success. It dropped on August 14, 2024 with better performance, sharper reasoning, and even image generation taking direct aim at GBT4. At the same time, XAI introduced a smaller version called Gro 2 Mini for speedier replies and soon after they released the full Gro 2 model to everyone. Think about it. In around a year, XAI went from zero to legit Chad GBT rival. Sure, it didn't instantly set the world on fire. Chad GBT still dominates, but the bigger story is that the technology has gotten better, cheaper, and simpler to use. I tried that image generation myself, and it's surprisingly solid. basically on par with other popular image generators. And we also have one built right into Geek Academy. I just open the AI art studio and type / imagine in the chat. And all Geek Academy members can generate images this way. Of course, our main focus is education, exclusive prompting guides, toll collections, and video lessons and AI that you won't find on YouTube. Our guides come as posts with prompts and explanations or as handy PDF documents you can download and keep forever. We are basically packing Geek Academy with everything you need to go from a total beginner to pro. Plus, we have special built-in tools to help you find exactly the right AI for whatever you need. Just call them out in the chat. You can ask for app recommendations or even have a full conversation if you like. Our goal is to make Geek Academy the biggest and best online learning hub for AI enthusiasts. If that sounds like your kind of place, just click the link below and join in. Now, with Grog 3, we're seeing a major leap toward a model that can do it all while staying objective and unbiased. Sure, it still leans a bit to the left, but it's a lot more neutral than plenty of other models out there. The only comparable AI shakeup might be Deepseek since it's also cheap, it's fast, and easy to scale. But let's be honest, Elon beat the Chinese to the

punch on this one. Right in the middle of all that OpenAI drama, Elon introduced Neurolink, Brain Implant. It's not strictly an AI project, but it's closely connected to AI, so it fits into the bigger story. So, what is Neurolink in a nutshell? It's a tiny device often called the Link that uses super thin electrodes to read and stimulate brain signals. In the short term, it aims to help people who are paralyzed control computers and prosthetics and maybe even restore sight or hearing. But Elon's bigger goal is to have Neurolink act as a direct link between our brains and advanced AI systems. Back in 2019, Elon said, "We will not be able to be smarter than a digital supercomput. So, we must create a high bandwidth interface to the brain if we want to go along for the ride. " Basically, with a high-speed neural link, you could send prompts straight from your thoughts and get immediate AI powered insights. This could seriously boost human thinking by mixing human brain power with AI analysis. On top of all that, all the brain data once it's anonymized and scaled up could give AI systems a big new source of info about how our neurons work. Though the timeline is still unclear, neural link might create a huge indirect impact on AI. The big thing is the overall concept and vision, even if we're not sure how

long it will actually take. All this makes you wonder, so is there a Tesla AI in the works? Well, it depends on what we call Tesla AI. In some ways, Grock already is Tesla AI, and I'm pretty sure will end up being the base for Elon's all-in-one AI down the road. But let's look at some real clues because there are plenty of signs this AI might be happening. One, hardware already exists. Tesla cars have been using some form of self-driving for years. And to train that AI, Elon's team built a supercomput called Dojo. It runs on in-house D1 chips made specifically for neural network training. That means Dojo isn't using the usual Nvidia GPUs like everyone else. Instead, Tesla developed its own AI focused chips, kind of like what Nvidia does, but all in Tesla's terms. Experts are split on Dojo. Some say it won't shake up the supercomputer world while others think it's a massive leap forward. Either way, Dojo is pushing hardware in new directions, which definitely plays a part in advance in AI. Two, advanced vision for full self-driving. Image recognition is huge for self-driving AI that Tesla wants to stick in all its cars. The car has to process hundreds of images a second, spotting road signs, other vehicles, pedestrians, traffic lights, and it needs to do this in real time, predicting paths and outcomes quickly. That's a seriously impressive AI challenge. Might not be as broad as Grock, but it's super focused and fine-tuned. These specialized AIs are valuable right now and building a high performance narrow AI and custom chips is a big hint that Tesla could one day create a fully independent AI system of its own. Three robots are in the works. Elon has mentioned Optimus project aimed at building a multi-purpose robot that in his words will be able to do anything you think of. Obviously for a robot to run on its own needs tailored AI. Elon said training a humanoid robot for general tasks will take about 10 times the AI effort of car self-driving and dojo is already working on a model for that. Sure, the robot failed future may be far off, but there's no denying that over time these robots could become platforms for Tesla AI. These robots would almost certainly be able to talk and respond to us naturally. Picture something like going to a website, turn on your mic, and just chatting. kind of like talking to a digital version of yourself. You could chat with me live, ask questions about Geek Academy or Galls to learn materials we have and more. There is a talk to AI master channel with a link where you can actually do this right now. It's a really cool way to learn. Taken together, these three basic examples already hint what Tesla AI might become. Something that feels like an AGI with vision, decision-making, and total independence from other platforms. But of course, like I mentioned, Grock is the biggest clue. Both the speed of its development and Elon's vision for it, not to mention its potential advantages over other models suggest we're looking

at a core technology that might power Elon's grand AI plans. What does Gro 3 do differently? Sure, under the hood, Grock feels like your typical LM like Chat GBT, but it actually uses a unique approach to solving problems. First glance, Grog 3 might look like just another upgrade from Grog 2, but it's actually a huge leap forward. For one thing, it's way more powerful. According to benchmarks from OpenAI, Grog 3 beats the competition on accuracy, speed, and even reasoning. Yep, that includes OpenAI's own OE models. Pretty impressive, right? Second, Grog 3's multimodality got a big boost. So now it not only sees better, it also creates images on a whole new level. I've looked at some of Rock 3's visuals and they are as good as what you get from the top names out there. I wouldn't be shocked if some Tesla's self-driving image processing magic ended up here, at least in how Grock handles and analyzes pictures. Then there is Deep Research, which is honestly fun to play around with. It's a lot like Chad GBT's deep search. So, there is not much more to add. Same idea with builtin and improved web search. But what really sets Grock 3 apart is how it looks for truth first instead of just spitting out any old response. Let's face it, a lot of AIS will throw out answers even if they are unsure. Grog 3 though aims to be factual and straightforward, even if it might be a little blunt. Elon wants it to handle questions that other AIs might avoid, still staying legal. Of course, I'm pretty sure Grog 3 with its enhanced multimodality will end up being the foundation for Tesla AI. It's already got multiple modes, strong reasoning, lots of cutting edge tech built in. All that's left is to scale it up until it hits AGI. And judging by how Grog jumped from a basic model to a real chat GBT rival in under 2 years, I'd say Elon stands a solid chance of beating Sam to the AGI finish line. And guys, let's not pretend it isn't weird how Elon Musk and Donald Trump go back and forth in AI. Both men treat the tech like a good

thing and look for ways to develop it, though some of their moves do seem a bit suspect. The president sees AI mostly as a business tool, something America should lead in. And he tends to take a more libertarian view and how it's regulated. Elon, on the other hand, sees AI as a major threat to humanity and keeps asking for stricter control over it. These two viewpoints seem to clash in the White House itself. After the $500 billion project Stargate project got the green light, Elon went off on X with harsh criticism. Stargate is supposed to build a complete AI framework in the US with partners like OpenAI, Soft Bank, and Oracle. And the second you hear OpenAI, you know Elon won't be thrilled about it. He wrote they didn't actually have the money. Soft Bank has well under $10 billion secured. I have that and good authority. Then Donald Trump said Elon Musk only reacted that way because he hates one of the participants obviously meaning AI. So we have to ask is Elon's reaction about money or is it about his principles? Meanwhile, Elon also wants to use AI to boost government efficiency which makes some folks uneasy mainly because of how he hopes to use it. According to the Washington Post, Elon's group has started running sensitive government data through AI programs to analyze spending and figure out what can be cut. His Doge team told the post that the ultimate goal is to replace the human workforce with machines. So again, what's the real angle here? This situation can change fast and we make sure to post all the updates in AI weekly updates and more detailed analytics and AI trends reports on Geek Academy. So, go check them out when you have time. It all begs the question, where is AI headed under this administration? Some experts hope Musk can convince Trump to treat super intelligent AI more like a national security risk that needs regulation rather than just another industry to deregulate. That might lead Musk to slow down the AI race, which could help keep things safer. AI is still so new that even its creators aren't always sure what it can do. So having proper rules and oversight, even if it slows progress, might cut down on the risks. But there is no denying something big is cooking. And we might see another post in X soon that gives us another clue about the mighty Tesla AI. From what we already know about Elon's companies and the latest AI trends, we can guess a lot about what this AGI might do. And honestly, things are looking pretty

clear. Not long ago, the Wall Street Journal reported that Tesla might license XAI's models to power self-driving, meaning XAI could take a share of Tesla's future revenue. That would include using Grog for an in-car voice assistant and the Optimus robot software. Exactly what we've been talking about. Elon even confirmed that XAI engineers have helped speed up Tesla's push for unsupervised full self-driving and that XAI will be helpful in advancing full self-driving and in building up the new Tesla data center. To me, that sounds like a plan to merge both companies tech. We don't have a full Grog guide yet, but do you want one? There's a poll at Geek Academy. Go vote and let us know. Meanwhile, XAI seems to be gearing up for a huge AI project by building massive data centers, one in Memphis, nicknamed Colossus, and another in Atlanta, to train and run its models. In Atlanta, XAI and X together shelled out for around 12,000 Nvidia GPUs, about $700 million to create an exoscale AI supercluster. Memphis Colossus is even bigger. Elon calls it the largest data center in the world. He's hinted that the nextG cluster might be five times more powerful. Sure, these aren't using Tesla's own chips, probably due to supply issues, but you don't build something five times bigger, unless you want a model that's five times smarter, maybe even an AGI. All this lines up with Elon's timeline for AGI and ASI in 2024. He said, "AGI might be here by 2025 or 2026. " And added, "If you define AGI as smarter than the smartest human, I think it's probably next year, within 2 years. " He didn't say it would be Grog 4 or Tesla AI, just that we're headed in that direction. A safer bet puts full AGI closer to 2029. And we know Elon can be pretty optimistic. But hey, those giant data centers aren't just for show. One thing's for sure, Elon doesn't give up. He's not stopping until Tesla AI or X AI or whatever he decides to call it becomes real. Unless politics take a big turn, he's got a decent shot at getting there. Of course, Open AAI is still way ahead, but I'm really hoping we'll see them get some serious competition soon. And, you know, we'll cover it the second it happens. So, if you haven't subscribed yet, be sure to do that. Thanks for tuning in and I'll catch you in the next one.