Farzad Mesbahi jumped into his Cybertruck and sped downtown, hunting for a glimpse of a driverless Model Y Juniper. Clips of the vehicle—turning west onto James Street from South Congress Avenue—began circulating online just a day after Austin officially listed Elon Musk’s company as a technology tester.

While Mesbahi, himself a former Tesla employee, gave up after a half-hour in the heat, he agreed with many bulls that the day would prove historic.

“This has been part of the [investment] thesis for a really, really long time,” he later told a podcast, explaining why he livestreamed his search for the elusive black crossover with the word “Robotaxi” emblazoned on the door. 

The widely-shared nine-second clip already has over 15 million views on Musk’s own X platform.

When asked by a fan “drooling over that one tiny clip” when public rides would start, the Tesla CEO also finally provided the first clear launch date.

“Tentatively June 22,” he wrote. “We are being super paranoid about safety, so the date could shift.” 

But if it does end up being postponed, it likely won’t be by much. The first Tesla that drives itself from the factory to a customer’s house, Musk stated, would be June 28—coinciding with the entrepreneur’s 54th birthday (how far a drive that would be, he notably left out, since Teslas cannot recharge themselves).

Wedbush tech analyst praised Musk for shrewdly timing his “apology tour with Trump” so as not to grab headlines away from the Austin pilot launch or otherwise risk the rollout. 

“We believe Tesla could reach a $2 trillion market cap by the end of 2026 in a bull case scenario,” wrote Ives on Wednesday. That would mean the stock price doubles in the next 18 months.

Waymo may have the lead, but Tesla can out-scale its competitors

For years Tesla has claimed it was on the cusp of solving unsupervised full self-driving, or FSD. When the company first launched FSD in late 2016 using a staged video it has since taken down, Tesla claimed the technology was already there—all it needed was to validate the software.

Six years ago, Musk already felt comfortable enough to predict he would have 1 million robotaxis on the road by 2020. 

Since then Tesla has undergone two full hardware upgrades to its AI inference computer, now in its fourth generation. Its FSD software is on its 13th iteration, not counting the myriad minor releases in between. In the meantime, it scrapped its previous approach of painstakingly coding commands in C++ language to adopt an AI-only approach and trained on its own Nvidia cluster, dubbed Cortex.

It’s important to note that Tesla is certainly not the first to launch a robotaxi pilot—Waymo has already graduated past testing to deploy its fleet commercially in several selected cities.

The significance then isn’t so much that Tesla may have potentially solved autonomous driving at long last—although that would be a major accomplishment in its own right—it’s Tesla’s ability to scale its fleet almost instantaneously that makes the launch so important. 

A decade in the making. Extremely proud of Tesla. This will save many lives, a lot of money, and a ton of time. https://t.co/BPFIlF3o8q

— Rob Maurer (@TeslaPodcast) June 10, 2025

Unlike competitors, Tesla doesn’t rely on multiple, expensive lidars and a host of radars that adorn each Waymo or Zoox robotaxi to precisely scan their environments in real time using laser beams and sound waves, in addition to other sensors.

Instead it relies solely on artificial intelligence to make sense of the two-dimensional images provided by its comparatively affordable cameras as eyes. Critics have long argued this is a high-risk gamble, however, since there’s no guaranteeing regulators would green-light robotaxis that lack a sensor suite as backup should the cameras fail. 

“That car you can buy right now for like 38 grand,” Mesbahi said about the Austin robotaxi vehicle captured in the video, “and that car is doing self-driving.”

Musk’s “ChatGPT moment” may soon be here

In fact, it’s not just the newer, refreshed Model Y that should be capable of FSD. Once Musks feel it’s safe enough, Tesla could push out its latest version of the software overnight to the hundreds of thousands of cars already on U.S. roads, so long as they are equipped with its fourth-gen AI inference computer commonly known as HW4. 

That’s because of a prescient decision made many years ago by the CEO to buck the usual industry practice of only equipping hardware a customer orders on purchase. 

Instead he had Tesla install these computers on every single vehicle that rolls off the factory line—whether a customer wanted it or not. While this cost Tesla extra, it would ensure older vehicles would be updatable and upgradable for the day when unsupervised full self-driving finally rolled out. Musk has called this Tesla’s “ChatGPT moment.”

Just to remove any final doubt he might be hiding anything, Musk explicitly stated that the Model Y robotaxi pictured in the clip was standard-issue and did not feature any concealed hardware.

“These are unmodified Tesla cars coming straight from the factory,” Musk said.