Investor's Corner
Tesla to focus on “narrow AI” and neural maps in next-gen Autopilot
Tesla reported Q2 financials yesterday which lacked new actionable information for Tesla traders, but provided some unexpected details on one topic, autonomous driving.
Neel N. Mehta of Morgan Stanley was first to ask for a “an update on Tesla’s proprietary mapping initiatives.”
Elon initially responded that “I think we would prefer to be confidential in that regard,” but Elon being Elon further explained that “what we’ve said thus far is that there’s need to have much higher definition maps than currently exists anywhere in the world in order to have full autonomy. And we’re in the process of building those and I think making good progress.”
Later on in the call, James J. Albertine of Consumer Edge Research asked to “understand in more detail I think how you [Elon] plan to get to fully autonomous.” Elon finally relented and said “Well, again, major product announcements are not – I shouldn’t do those on an earnings call, obviously. And all I’d say is that full autonomy is going to come a hell of a lot faster than anyone thinks it will. And I think what we’ve got under development is going to blow people’s minds. It blows my mind, so.”
Elon was not done. Colin Rusch of Oppenheimer went back on the subject and asked “how you guys are going to approach that functionality going forward with the driver assist in the autonomous driving push going forward?” Elon answered that “I think we’ll have a more significant announcement on that later. So it’s not really – earnings call is not the right time for that except that it will be a Tesla solution, internal solution.”
The analysts were not done with the topic. Brad Erickson of Pacific Crest Securities pounded on asking “Just had a quick follow up, I guess, on something that’s been asked a couple of times; take another run at it. I guess given that you’re obviously no longer working with this key supplier [Mobileye] around full autonomy. What are the major hurdles that you see for Tesla here to overcome to get to full autonomy? Is it just a case of software development, lots more miles driven and basically getting the right people in place? Any color on sort of some of the key challenges you’re facing and where you’re particularly focused for delivering full autonomy at some point?”
“It’s exciting, it blows me away, the progress we’re making.” – Elon Musk
Elon finally gave some more details: “Well, full autonomy is really a software limitation. I mean the hardware is just to create full autonomy, so it’s really about developing advanced, narrow AI for the car to operate on. I want to emphasize narrow AI, it’s like not going to take over the world, but it needs to be really good at driving a car. So increasingly sophisticated neural maps that can operate in reasonably sized computers in the car. That’s our focus. I’m very optimistic about this. It’s exciting, it blows me away, the progress we’re making. So I think if I’m this close to it and it’s blowing me away, it’s really going to blow other people away when they see it for the first time.”
-
According to Wikipedia, “narrow AI,” also known as “weak AI”, “defines non-sentient computer intelligence or AI that is focused on one narrow task. Weak AI is defined in contrast to either “strong AI” (a machine with consciousness, sentience and mind) or “artificial general intelligence” (a machine with the ability to apply intelligence to any problem, rather than just one specific problem).” Wikipedia cites Apple’s Siri as an example of a “narrow AI”.
Combining all of Elon’s responses, we can expect a major [significant] announcement regarding an internal Tesla solution to autonomous driving, which will involve a combination of “narrow AI” software implementing sophisticated neural maps, and much higher definition maps but likely no major new hardware. So much for being “confidential.”
This is great news to current Tesla owners because it is a “software” solution: current Model S, Model X and eventual Model 3 owners may be capable of full autonomous driving.
If Elon thinks that all of this “is going to blow other people away”, it probably will.
Photo credit: Garth Woodworth
-
Cybertruck
Tesla Cybertruck production snaps back after ugly supplier fight
Cybertrucks are piling up again at Giga Texas after Tesla’s court win against a parts supplier.
Cybertruck production at Giga Texas is showing its first visible recovery since Tesla sued a supplier last month over withheld manufacturing tooling.
Aerial observer Joe Tegtmeyer flew over the Austin factory Wednesday morning and counted roughly 100 or more Cybertrucks filling the outbound lot, a sharp jump from the thin numbers seen in recent weeks. The flyover came a day after a judge granted Tesla a temporary restraining order against Angstrom Automotive Group, the parts supplier at the center of the dispute.
Tesla filed an emergency lawsuit in late July after Angstrom told the automaker it planned to close the Troy, Texas facility where Tesla’s die-cast tools, trim dies and other Cybertruck stamping equipment were housed. According to Tesla’s complaint, a shipment of 700 finished parts never left the building, and when Tesla sent representatives to retrieve its equipment, accompanied by law enforcement, they were turned away. Angstrom allegedly then asked for an extra $250,000 a week to keep operating, which Tesla’s filing described as holding its own property for ransom.
TESLA: U.S. District Judge Christopher R. Wolfe of the U.S. District Court for the Western District of Texas, Waco Division granted Tesla a Temporary Restraining Order and Writ of Replevin in its dispute with Angstrom Automotive (Case No. 6:26-cv-00477).
The order authorizes… https://t.co/E1DKcQSxMn pic.twitter.com/LR8aAiV2Og
— S.E. Robinson, Jr. (@SERobinsonJr) August 5, 2026
-
The restraining order gives Tesla immediate right of entry to Angstrom’s facility to recover the tooling. It is temporary, with a fuller hearing still to come, but the speed of Wednesday’s rebound suggests the Angstrom shortage was indeed the main bottleneck limiting Cybertruck output. Outbound lot counts are an imperfect measure of actual production, since finished trucks can sit for days before shipping, but a lot that full after a lean stretch is a meaningful signal.
Cybertruck output at Giga Texas has fluctuated all year as Tesla worked through supply issues and introduced new trims, including a cheaper Dual Motor AWD version that drew strong early demand.
Venture capitalist Chamath Palihapitiya has cautioned investors shorting SpaceX shares, drawing a direct parallel to the intense short-selling pressure Tesla faced in its early public years.
Responding to reports of elevated short interest in the newly public rocket, satellite, and AI company, Palihapitiya noted that similar dynamics played out with Tesla, where aggressive short sellers ultimately “went broke.”
SpaceX (NASDAQ: SPCX) went public on June 12, 2026, in the largest IPO on record, pricing at $135 per share. Shares quickly surged to an all-time high of $225.64 just days later, briefly implying a valuation exceeding $2 trillion. The stock has since retreated sharply amid valuation concerns, lockup expiration fears, and broader market dynamics.
By early August, it traded near $108–$125, representing a roughly 50 percent decline from the peak and bringing the market capitalization closer to the $1.5–1.7 trillion range. On August 4, shares closed up more than 9 percent at $125.33 ahead of earnings before facing pressure in after-hours and premarket trading.
Short interest has climbed dramatically. According to S3 Partners data widely cited in market reports, short positions reached approximately 219.3 million shares by late July, about 34 percent of the limited public float of roughly 640 million shares, and represented a notional value of around $24.6 billion.
Utilization of shares available to borrow hit 95 percent, with borrow fees rising. This level of shorting exceeded the dollar value of short bets against Tesla at the time and built rapidly ahead of two catalysts: the company’s first post-IPO earnings and an August 6 lockup expiration that could free up to 911.5 million additional shares.
-
CEO Elon Musk has issued warnings of his own. In mid-July, as short interest approached one-third of the float, he posted that “the survival probability of firms who maintain a significant short position in SpaceX over time is very low,” reiterating his view that the company could ultimately be worth more than Earth if it achieves its goals.
On August 4, just before earnings, Musk responded to the latest short-interest data by saying, “I try to warn them, but they just double down.”
SpaceX delivered its first quarterly results as a public company after the close on August 4. Second-quarter revenue rose 92 percent year-over-year to $7.8 billion, beating consensus estimates near $6.8–6.9 billion.
The net loss narrowed to $541 million, or 9 cents per share, better than the roughly 23–24 cent loss expected. Starlink/connectivity contributed about $4.3 billion (up 66 percent), while the AI business generated $2.6 billion (up roughly 250 percent). Capital expenditures were heavy at $18.4 billion, largely tied to AI infrastructure. Management projected a $100 billion annualized revenue run rate by year-end 2026 and outlined a path toward $1 trillion in annual revenue by 2030.
The combination of Chamath’s historical reminder, Musk’s repeated alerts, and the company’s ambitious growth targets underscores the high-stakes debate surrounding SPCX. Short sellers are positioned for near-term supply pressure from the lockup, while long-term bulls point to Starlink scale, Starship progress, and AI compute expansion as reasons the bears may ultimately face the same fate as many early Tesla skeptics.
Investor's Corner
SpaceX and Nvidia team up on Musk’s orbital AI bet
SpaceX revealed a new Nvidia satellite partnership, then Musk pledged an exclusive Nvidia hardware commitment.
SpaceX and Nvidia are now working together on the hardware that will power Musk’s orbital data center ambitions. SpaceX announced on X on Tuesday that it is partnering with Nvidia to design the compute payload for Starmind AI1, the first satellite in a planned constellation built to run AI workloads directly in orbit. Each Starmind satellite will carry Nvidia’s Rubin GPUs and Vera CPUs, according to the post, which included renderings of the payload design.
The announcement landed hours before SpaceX’s first earnings call as a public company, where Musk went further, saying the company has committed to building its AI infrastructure exclusively on Nvidia hardware. “We think the Vera Rubin architecture is the best architecture. We think it’s the best AI computer, and we greatly value our close cooperation and partnership on many levels with Nvidia,” Musk told investors on the call,. “So we’re exclusive to Nvidia.”
Musk said SpaceX plans to deploy Nvidia’s Vera Rubin NVL72 rackscale system, codenamed Kyber, both on the ground and in space. He set a target of 2 gigawatts of compute capacity online by the end of this year, scaling to roughly 10 gigawatts by the end of 2027.
SpaceX’s newest Starmind will make earth data centers obsolete
Starmind has been in development since Musk confirmed the name in June, following an xAI trademark filing that tipped off the project before SpaceX made it official. The idea is massive in scope and instead of moving data down to ground based servers, satellites equipped with onboard processors and large solar arrays would compute AI workloads in orbit and beam results back to Earth. SpaceX has already filed with the FCC for a constellation of up to one million satellites to support the effort, citing constant solar power and the absence of zoning restrictions as advantages over terrestrial data centers.
The Nvidia exclusivity marks a shift in tone from just two weeks ago, when Musk was busy knocking down a report that SpaceX had ordered $52 billion worth of Nvidia GPUs through Foxconn, calling it fake news at the time. The dollar figure in that rumor may have been wrong, but the underlying direction seems correct. SpaceX’s AI division already leases Colossus compute capacity to Anthropic and Google, and Tuesday’s earnings report showed AI revenue climbing sharply as those deals ramp up.
Nvidia shares rose roughly 3% in Tuesday trading on the news, while SpaceX stock climbed nearly 9% during the day before giving back gains after hours as investors digested the earnings report’s capital spending figures.
-
Tesla Cybertruck production snaps back after ugly supplier fight
Space finally faced the people living next to its next Terafab mega-project
