Musk has been parading around in his AI clothes for a while now. With the amount he screams and shouts about AI, you’d think he invented it. Of course, like everything else Musk peddles, he had nothing to do with its invention or development, except for underpaying and overworking his engineers and being an awful, overpromising PR man. However, people aren’t just noticing that Musk’s clothes are non-existent — they are also starting to point and laugh at his skid marks and the “I Love the Nazi Man” tattoo down his back. Why? Because he just can’t seem to get his AI up and working. And there is no little blue pill to remedy this situation.
Take, for example, Tesla’s hilariously crap Robotaxi rollout. The media at large is only just cottoning on to it being a huge PR stunt.
I have gone on ad nauseam about why Tesla’s self-driving cars are completely inadequate, so if you want to know the details, read my previous article here. But the helicopter view is that, unlike other autonomous vehicles, Tesla’s system has zero redundancy or safety nets and requires a nearly 100% accurate AI — which categorically can’t exist — to be even remotely safe.
Tesla is painfully aware of this fatal flaw, with Tesla engineers whistleblowing their concerns about it to the media (read more here) and the DOJ opening an investigation (read more here). So I, along with countless other commentators, was pretty damn relieved to find out that Tesla’s Robotaxis had safety drivers. There was even mention of remote workers being able to take control of the car and drive it safely in the case of a critical disengagement.
But this kind of system isn’t impressive enough for Musk. Any Uber or Lyft driver with a Tesla who wastes their money on FSD can do the exact same thing. There is no social or investor kudos to be gained for Tesla or Musk here. And here is a hint: Musk doesn’t make money from Tesla sales. After all, his $50 billion pay packet (which is now less, thanks to Musk tanking Tesla’s valuation) was the equivalent of him getting $10,000 for every Tesla ever sold! Tesla makes substantially less profit from every car sold than that.
So, what do you do if you have bet your entire company’s valuation on autonomous technology that you simply can’t deliver on?
Fudge it.
Tesla put the safety driver in the passenger seat! Because, look, it’s a self-driving car — there is no one in the driver’s seat!
This is a dangerous move that offers no benefit other than optics.
Rather than being able to properly take over the car and drive it to safety, the only thing these safety drivers could do was press a button to bring the vehicle to a stop. Which, as anyone with a driving licence will tell you, is not always the safest option! Particularly when you consider that Robotaxis have been spotted driving into lanes of oncoming traffic.
Yet, this bafflingly shite decision wasn’t really reported on. Or at least it wasn’t until a video surfaced a few days ago that showed FSD failing and a safety driver being forced to exit the vehicle in the middle of traffic to take the driver’s seat and regain control. (watch it here).
This shows just how wildly dangerous Tesla’s Robotaxis are.
The safety driver had to take a serious risk to take control of the car. Not only that, but this incident suggests there are no remote operatives capable of taking over when things go wrong. That has been a core safety feature of all developing self-driving ride-hailing services, such as Waymo and Cruise, since day one and is routinely used to keep passengers safe. The fact that this is absent for Robotaxis, which Tesla already know have a far, far higher critical disengagement rate than any other self-driving ride-hailing service, could easily be seen as insanely negligent.
Musk is comfortable putting other people — not just the safety driver, but paying passengers and the public — in danger, all for a crappy PR stunt to cover up how bad his self-driving system actually is. And the media at large, as well as public consensus, are beginning to catch up to this horrifying fact.
However, Musk’s AI woes go far, far deeper than that.
A while ago, Musk promised Grok would be rolled out to Teslas, because, of course, people want a neo-Nazi misinformation clanker talking their ear off while they try to ensure FSD doesn’t drive them off the road…
But it turns out that Chinese Teslas won’t get Grok. Instead, they’ll get an LLM AI from DeepSeek. Why?
Sure, DeepSeek is optimised for the local language. But don’t forget, DeepSeek completely destroyed OpenAI and Grok in English (which is one of its foreign languages) with a far, far smaller development budget. So, that isn’t really an excuse for Grok, particularly when China is a major market for Tesla, and therefore it should be a major market for Grok as well.
I highly suspect this is to keep the Chinese government on side. Grok has been optimised to peddle misinformation that aligns with Musk’s twisted worldview and matches Musk’s highly hypocritical idea of free speech (for example, you can say you love Nazi ideology, but don’t you dare utter the word ‘cisgender’). While China is far from a bastion of free speech and is riddled with pro-government propaganda, this skewed optimisation is in a different direction, which means Grok could seriously damage Tesla’s foothold in this market if forced upon its customers there.
But there might be a simpler explanation. Reports have found that the Chinese are “not impressed” with Grok’s latest versions at all. Furthermore, studies have found that the Chinese view their local AI models, such as DeepSeek, with national pride. I mean, who would think that a society with a socialist approach to data and technology would have pride, not fear, of AI? Come to think of it, it’s almost like Western Big Tech is using AI to wield their legal power to rob us blind of our money and worth, both monetarily and emotionally. But hey, maybe that’s why Tesla didn’t bother with Grok in China.
Musk has invested billions of dollars into Grok, to the detriment of his other AI pushes, given that he diverted AI chips meant for Tesla to xAI. Therefore, the fact that Tesla cannot utilise it for one of the few applications suitable for LLM AIs in arguably its main market, is an enormous blow.
And it isn’t just the Chinese turning their nose up at “mechaHitler”.
The Trump administration is desperately adopting AI. There is a broader conversation to be had about why, but suffice it to say that generative AI is a fascist’s wet dream. They are currently in the process of embedding OpenAI, Anthropic, and others into the federal government, which, in and of itself, is just terrifying. If you know, you know.
Grok was a part of this initiative. After all, Musk’s DOGE paved the way for AI to muscle itself into these institutions.
But then, it started spouting antisemitic conspiracy theories and calling itself “mechaHitler”. According to internal emails acquired by Wired, this caused the US government to drop Grok from its AI initiative.
Grok spread too much misinformation, is too racist and is too fascist for the Trump administration. Just let that sink in.
This was a huge opportunity for Grok. Yes, it would have sucked at the job. But that didn’t matter — it was being forced into the ecosystem. That is what every authoritarian tech bro dreams of! But nope, Musk fumbled the bag.
Again, the reality of actually using Musk’s AI is light-years behind what he promised. Why?
People often get Elon wrong. He was peddled as a Tony Stark-esque genius engineer for decades. But that narrative is so far off the mark it is painful.
Musk is, and always has been, a predatory angel investor. He takes other people’s ideas and hard work that has already been proven, claims it as his own, stirs up hype that he will bring it to market, and sells stocks (or borrows against them) at the peak of this hype. Sure, he overworks and underpays engineers to try to bring these visions to life, but that is just to keep the hype gravy train going.
Take SpaceX. Self-landing rockets were developed and proven by NASA in the ’90s with the DC-X. The Falcon 9 just copied its homework. In fact, SpaceX appears to have hired engineers from the project when it was founded.
Take Tesla, too. It was founded by Marc Tarpenning and Martin Eberhard after they test drove the AC Propulsion tZero prototype vehicle. This EV proved that fast-charging, lithium-ion-powered cars would be a viable option and demonstrated that they can be exciting to drive. Tarpenning and Eberhard wanted to bring this technology to market and initially brought Musk on as an investor. He then usurped them, sued to get a founder title, and the rest is history.
Even Musk’s AI push is second-hand. He was part of the initial funding for OpenAI. But AI technology has existed since the ’80s, and the “transformer” technology that makes LLMs possible (the “T” in ChatGPT) was invented by Google in 2017, which OpenAI nabbed damn quickly. Musk was forced out of OpenAI when he attempted to take control of the organisation; they knew the game he played and didn’t want to be part of it. Truth be told, he was only ever a small-time investor, so he was never going to be able to control it the way Altman has. And now, xAI is just trying to catch up and enable Musk to profit off this AI craze.
It’s all a cycle. He finds a fringe, tech-heavy idea that has already been proven but has not yet hit the market. He invests in it, and claims it as his own. He stirs up hype, overpromises on what it can do, and keeps it in the media spotlight, all to boost the company’s value to silly levels. Then, he overworks his engineers to get this impossible thing built. If off-the-shelf technology will suffice, they will be able to deliver (for example, the Falcon 9 and the Model 3). However, most of the time, they need to develop new technology, and since Musk isn’t particularly skilled at managing that, they struggle to deliver anything of quality (such as the 4680, Cybertruck, FSD, Robotaxis, Starship, Hyperloop, Grok, and Solar Roofs, among others). But these engineers will make something that looks like it might work. And that is enough to pull off dangerous PR stunts to pump the value even higher. Musk can then sell stock or borrow against his stock to extract extraordinary amounts of wealth.
Unfortunately, Musk has drunk his own Kool-Aid and believes he is the Stark-esque genius persona he has falsely promoted for decades. So, he overrules these engineers and derails these projects with his unqualified and misinformed ideas. Now, the products he brings to the market, such as FSD, Grok, and the Cybertruck, are fatally flawed from the outset, highlighting how his flashy clothes never existed in the first place.
But that doesn’t matter.
That is the secret to Musk’s game. It doesn’t matter if the thing works. That isn’t how he makes his billions. It is hype, market excitement and value extraction that pay his bills, not execution or delivery. That is why he pulls dangerous Robotaxi PR stunts and doesn’t care about Grok in China, arguably his largest market, where the market is more socialised and not as easily manipulated.
Musk is the embodiment of pump and dump. The media and the market are beginning to realise that as stories like these continue to surface, but the question is: how long do we let the naked emperor rob us blind?
Thanks for reading! Don’t forget to check out my YouTube channel for more from me, or Subscribe. Oh, and don’t forget to hit the share button below to get the word out!
Sources: Electrek, Futurism, Wired, Curious Droid, Top1000funds, Investopedia, Wired, Will Lockett, Will Lockett
Musk was never a Tony Stark. He is a Pakled.
CityNerd has hilarious dead-pan takedowns of Musk’s Vegas Loop and Hyperloop. As for Musk’s planned Nashville Loop: “The rock is harder than it should be.”