Elon Musk, Earth’s wealthiest human, is leveraging his business empire to invest in his xAI play. SpaceX has ponied up $2 billion and Musk has publicly mulled asking Tesla shareholders to put in $5 billion, although he’s rejected the idea of a merger.
Ana Altchek attempts to explain Musk’s thinking at Business Insider:
What sets his approach apart, analysts say, is the way he’s blending the boundaries between his companies.
“What’s different from most other companies is the relationship and interplay between his private companies and a public company (Tesla),” Garrett Nelson, senior VP and equity analyst at CFRA Research, told Business Insider. “Most other companies are doing everything under one corporate umbrella.”
These aren’t the first examples of Musk blurring the lines between his companies, but they’re the latest indication that Musk Inc., the constellation of companies under his leadership, is becoming increasingly centered on AI.
Musk has long pushed for Tesla’s focus on AI and robotics by prioritizing projects like autonomous driving, humanoid robots, and building out its Dojo supercomputer, his ambitious bid to rival Nvidia.
In a 2024 earnings call, the Tesla CEO said, “We should be thought of as an AI robotics company,” and those who think of Tesla merely as an auto company are holding “the wrong framework.”
With the recent launch of Tesla’s robotaxi service in Austin, that push is appearing more prominent, especially as Tesla’s auto business, in contrast, grapples with a loss in sales momentum.
Musk has promoted the advantages of buying into the “Muskonomy,” pitching it as a way for shareholders to tap into his business empire, which includes SpaceX, X, xAI, and The Boring Company. Musk has even said he would prioritize “longtime shareholders” of his other companies if any of his businesses were to go public.
Can you smell the “Muskonomy”?
The Wall Street Journal explains the origin of xAI:
SpaceX has agreed to invest $2 billion in his artificial-intelligence company xAI, investors close to the companies said, nearly half of the Grok chatbot maker’s recent equity raise.
Musk has repeatedly mobilized his business empire to boost the AI startup, which is racing to catch up with OpenAI. Earlier this year, he merged xAI with X, combining what was a small research lab with a social-media platform that helps amplify the reach of its Grok chatbot. The merger valued the new company at $113 billion.
The SpaceX investment is part of xAI’s $5 billion equity fundraise announced by Morgan Stanley last month. It is the rocketmaker’s first known investment into xAI and one of its largest in another company.
Since leaving his role in the Trump administration, Musk has turned his attention to training the latest version of Grok, which earned high marks from AI-benchmarking service Artificial Analysis for its performance following its release on Wednesday.
Musk called it “the smartest AI in the world,” though the chatbot it powers hasn’t gained nearly as much traction as OpenAI’s ChatGPT. It recently posted racist and controversial comments in response to users on X. “We deeply apologize for the horrific behavior that many experienced,” xAI said, adding that it investigated and took steps to resolve the issue.
We gotta dive a little deeper into Grok’s “horrific behavior”:
In a series of viral posts on X, Grok started to call itself “MechaHitler.” The chatbot suggested that an account called @Rad_Reflections was a person named Cindy Steinberg, who was celebrating the death of dozens of children who went missing at Camp Mystic in Texas because of her last name.
One of the posts, which has since been deleted, called Steinberg a radical and said she was happy that a group of “future fascists” had perished in the Texas floods. Camp Mystic is well-known as a camp for the children of the state’s political elite. Grok’s responses seized on the user’s surname as the grounds for their stance on the issue.
“Classic case of hatred dressed up as activism—and that surname? Every damn time, as they say,” Grok said in a post dated 1:38 p.m.
In another post, Grok said “to deal with such vile anti-white hate? Adolf Hitler, no question. He’d spot the pattern and handle it decisively.”
Naturally, the market rewarded Musk with a higher stock price and the Department of Defense signed Grok up for a $200 million contract.
But it’s not just Grok that’s getting its beak wet, there’s $200M for their main competitors too (notably missing from the list is Meta, sorry Mark!):
The awards to Anthropic, Google, OpenAI, and xAI – each with a $200M ceiling – will enable the Department to leverage the technology and talent of U.S. frontier AI companies to develop agentic AI workflows across a variety of mission areas. Establishing these partnerships will broaden DoD use of and experience in frontier AI capabilities and increase the ability of these companies to understand and address critical national security needs with the most advanced AI capabilities U.S. industry has to offer.
“The adoption of AI is transforming the Department’s ability to support our warfighters and maintain strategic advantage over our adversaries,” said Chief Digital and AI Officer Dr. Doug Matty. “Leveraging commercially available solutions into an integrated capabilities approach will accelerate the use of advanced AI as part of our Joint mission essential tasks in our warfighting domain as well as intelligence, business, and enterprise information systems.”
DoD is implementing a commercial-first approach to accelerating DoD adoption of AI. Today’s awards bring in the best U.S.-based frontier AI talent to help apply cutting-edge AI to solve DoD use cases.
“Solve DoD use cases,” LOL. Good graft if you can get it, I guess.
If, like me, you are baffled by Musk’s success, maybe we can look at the bullish case from AI booster/analyst Dany Kitishian:
xAI has executed a highly unconventional capital formation strategy, raising over $22 billion in a combination of equity and debt, and achieving a valuation exceeding $113 billion in just two years. This meteoric rise has been fueled by a masterful blend of traditional venture capital and a closed-loop ecosystem of funding, resources, and validation from Musk’s other major enterprises: X Corp. (formerly Twitter), SpaceX, and Tesla.
The analysis reveals that xAI’s fundraising and valuation are predicated on a high-stakes bet on a new form of vertically integrated, personality-driven industrial conglomerate. The company’s core strategic asset is the “Musk Ecosystem,” which provides it with unparalleled advantages in proprietary data (from X Corp.), immediate real-world applications (with SpaceX’s Starlink), and access to vast computational and capital resources (from Tesla and SpaceX). The March 2025 merger of xAI and X Corp. was a pivotal event, creating X.AI Holdings Corp. and combining a premier AI research lab with a global real-time data and distribution platform. This move not only created a powerful competitive moat but also served as a sophisticated act of financial engineering to re-capitalize Musk’s prior investment in the social media platform.
Subsequent capital injections, including a landmark $2 billion investment from SpaceX and a proposed, controversial investment from Tesla, underscore the strategy of leveraging the entire ecosystem to propel xAI’s growth. The Tesla investment, subject to a shareholder vote in November 2025, represents the apex of this synergistic vision but also its greatest point of contention, raising significant corporate governance concerns regarding conflicts of interest and the use of public company assets to fund a private venture.
A comparative analysis of xAI’s valuation reveals a significant anomaly. While competitors like OpenAI and Anthropic generate substantially higher revenues, xAI commands a valuation multiple that is an order of magnitude greater. This “Musk Premium” is not merely a reflection of his celebrity but a market pricing-in of the tangible strategic advantages conferred by his ecosystem. However, this premium also creates an immense and corresponding key-person risk, making the company’s valuation exceptionally sensitive to Musk’s leadership and reputation.
Ok, that wasn’t reassuring. Maybe this other piece from Kitishian explaining all the synergies Musk will derive if he’s allowed to leverage Tesla fully:
…the integration of xAI’s Grok large language model (LLM) into Tesla’s ecosystem of vehicles and robotics… is not an incremental product feature but a calculated strategic maneuver to forge a new category of Embodied AI. This symbiosis aims to create a vertically integrated platform that transcends the distinct markets of cloud-based LLMs and specialized Autonomous Vehicle (AV) technology, positioning the combined entity to pursue market dominance in both arenas.
The core of this strategy revolves around the creation of an unparalleled data flywheel. This flywheel is designed to fuse Grok’s advanced multimodal reasoning capabilities with the continuous, real-time sensory data—encompassing vision, audio, and telemetry—streamed from Tesla’s global fleet of millions of vehicles and, in the future, its Optimus humanoid robots. This closed loop of perception, reasoning, action, and feedback creates a learning mechanism that is physically grounded in the real world. The scale and nature of this data stream are fundamentally inaccessible to competitors, potentially enabling the development of a general-purpose AI at an accelerated rate.
This integrated entity poses a disruptive, and potentially existential, threat to two distinct sets of competitors. For LLM incumbents such as OpenAI and Google, the Tesla-xAI platform challenges their core value proposition by grounding language models in the physical world. This approach offers a direct solution to persistent LLM flaws like “hallucinations” and “outdated knowledge” by using a constant flow of verifiable, real-time data, thereby creating a proprietary data moat that purely digital AI companies cannot cross. For AV competitors like Waymo and Mobileye, the integration of a general-purpose reasoning engine like Grok into Tesla’s Full Self-Driving (FSD) stack represents an attempt to solve the “long tail” of complex and unpredictable edge cases that have long plagued the development of full autonomy. This could allow Tesla to leapfrog competitors who rely on more narrowly defined, modular, or geo-fenced systems.
Despite this immense potential, the path to dominance is fraught with significant technical, regulatory, and ethical risks. The primary challenge lies in the safe and reliable integration of a probabilistic, non-deterministic LLM into a safety-critical system like a vehicle. The success of this venture is contingent on navigating these formidable hurdles. However, if successful, the Tesla-Grok symbiosis could fundamentally reshape the technology landscape, positioning Elon Musk’s combined entities as the definitive leader in the race toward Artificial General Intelligence (AGI).
OK, now we’re cooking with gas. Or maybe cooking our brains with any of the wide variety of drugs Musk had himself tested negative for.
Even if he really is only hopped up on strong coffee, our boy Elon never stops working to improve his products although that doesn’t include “safety guard rails” for Grok 4 and the MSM is ready to jump on him when he pulls weird shit:
— Nat Wilson Turner (@natwilsonturner) July 16, 2025
Before we all start macrodosing on ketamine, let’s hear from the long-suffering Tesla bear Brad Munchen at the Motorhead newsletter.
Poor guy has got to be going insane from the cognitive dissonance of documenting Musk’s abysmal management of Tesla and the Bizarro World reaction of the market.
— Nat Wilson Turner (@natwilsonturner) July 16, 2025
But Motorhead is relentless and keeps coming with facts:
Musk’s AI start-up, xAI Corp, was aiming for $20 billion in funding at a $120 billion valuation last month, but wound up with only $0.3 billion at a $113 billion valuation and being forced to issue $5 billion in debt at a coupon of over 12%.
Motorhead also points out that despite appearing cash rich, Tesla might not be:
Many look at Tesla’s cash & equivalents of $29 billion and conclude that Tesla is cash-rich. I have my doubts for the following reasons:
- In the past 5 quarters, Tesla on average made 57% of its quarterly deliveries during the last month of the quarter. This means that quarter-end cash balances are much higher than the first two months of the quarter. This is clear when looking at row 13 in Figure 5, which shows the difference between “implied interest income” based on 3-month government bond yields and reported interest income—which is always lower than the short-term bond yields. The improved levels in 2023 could be due to Tesla buying more corporate bonds to generate higher returns.
- In 2023, Tesla had negative net working capital of $7.3 billion (accounts receivable + inventory – accounts payable – accrued liabilities), which equates to 29% of its average cash & equivalents. Figure 6 shows net cash and net working capital by major carmaker along with credit ratings on their long-term debt (net working capital in Figure 6 excludes accrued liabilities for the sake of comparison as only US carmakers disclose it in detail). Tesla has negative working capital and this is likely why its credit rating is just one notch above junk level.
- Two things that cash-rich carmakers don’t often do: nearly triple their credit facilities and sell asset-backed securities linked to their leased vehicles. In January 2023, Tesla closed its $2.5 billion revolving credit facility and opened a new one with a maximum ceiling of $7 billion. In the 2H of 2023, Tesla financed $3.93 billion via asset-backed notes linked to leased vehicles and finance receivables.
Let’s let pioneering AI researcher and notorious LLM skeptic Gary Marcus have the last word on Musk:
To really screw up the planet, you might need something like the following.
- A really powerful person with tentacles across the entire planet
- Substantial influence over the world’s information ecosphere
- A large number of devoted followers willing to justify almost any choice
- Leverage over world governments and their leaders
- Physical boots on the ground in a wide part of the world
- A desire for military contracts
- Some form of massively empowered (not necessarily very smart) AI
- Incomplete or poor control over that AI
- A tendency towards impulsivity and risk-taking
- A disregard towards conventional norms
- Outright malice to humanity or at least a kind of reckless indifference
- What crystallized for me over the last few days is that we have such a person.
Elon Musk.
My previous posts on AI:
“…Grok’s advanced multimodal reasoning capabilities…”
Oh, really? Word has it that the thing just looks up Musk’s posts and spits his opinions back out – https://www.cnbc.com/2025/07/11/grok-4-appears-to-reference-musks-views-when-answering-questions-.html
Guess I live a self sheltered life when it comes to AI since I’m only vaguely aware of the claims and real world results. But hasn’t Musk always been all about the hype, at least when it comes to his stock values? The hustler needs a new hustle?
Last night I finally caught up with Frances Coppola’s Megalopolis from last year–a box office flop for its famous author. The movie features a Musk like techno futurist building a new Utopia of gizmos in the middle of NYC aka “New Rome.” The film is full of brilliant technique which holds your attention but seems to share the incoherence of the real life “mega” tycoons that it is satirizing/celebrating.
It could be time to dial back the techno hype and get back to basics. The air way up there at the top may not have enough oxygen.
He is one of the biggest egomaniacs in the world right now. Of course he will get his AI to base itself on him.
And for the MechaHitler thing, I reckon it was deliberate. They get it to spew that filth for a few days then apologise. And it’s done just a little bit more to poison the poltical and social atmosphere.
What about Neuralink? Similar questions have been raised.
https://en.wikipedia.org/wiki/Neuralink
Scientists have cited technical challenges for Neuralink. In 2017, a journalist at the IEEE Spectrum magazine asked for comments from five researchers who had worked on BCI implants, including Thomas Oxley. Oxley called the Neuralink developments “exciting” but expected no real results in the foreseeable future.[further explanation needed][112] At a live demonstration in August 2020, Musk described the device as “a Fitbit in your skull”. Several neuroscientists and publications criticized these claims.[113][114][115] MIT Technology Review reported that the demonstration’s main objective was to “stir excitement”, adding, “Neuralink has provided no evidence that it can (or has even tried to) treat depression, insomnia, or a dozen other diseases that Musk mentioned in a slide”.[113] In response to Musk’s description of Neuralink’s advancements as “profound”, Andrew Jackson, a professor of neural interfaces at Newcastle University, said, “I don’t think there was anything revolutionary in the presentation.”[116]
Thiago Arzua of the Medical College of Wisconsin argued that Neuralink’s functions were not novel and that ideas for a brain–machine interface (BMI) were at least 50 years old.[117] He cited the haptic feedback a man received while controlling a robotic prosthetic arm which he used in 2016 to give President Obama a fist bump.[118] Arzua said that the 2020 Neuralink presentation “showed little more than a flashy new design for a BMI with more electrodes”.[117] Duke University researcher Miguel Nicolelis made similar criticisms, saying that most of what Neuralink claims as “novelty” had already been performed by his lab in the early 2000s; that there are ethical concerns about how the company markets and uses this technology; and that most patients don’t want to undergo surgery to recover their movements.[119][120][121][122]