Tesla CEO Elon Musk could have taken the stage at last night’s “We, Robot” event and put a lot of fears to rest.

He could have released comprehensive safety data for the company’s Full Self-Driving feature that showed real progress for the driver-assist feature, contradicting all the crowdsourced data that’s out there making FSD look truly awful.

He could have announced that the Cybercab, a sleek little two-seater with butterfly-wing doors, would be a geofenced, Level 4, fleet-owned vehicle, operating in a few select markets with impressive-looking margins.

He could have provided an ounce of detail about the Cybercab’s technology stack, including its sensors, vision system, and onboard processing power. And he could have shocked the industry and surprised many of his doubters by embracing lidar, the laser sensor that serves as a crucial redundant system for every other driverless vehicle on Earth.

But he did none of those things. Instead, he put on what arguably looked like a great show, complete with fake movie posters, a ton of delicious-looking food, and robot bartenders. And he fell back on the same old, tired promises of a fully autonomous vehicle that was “just two years away.”

We’ve been down this road before. Many times.

“Prototype hardware that works in a limited demo is cool, interesting, and fine to comment on,” Phil Koopman, an AV expert from Carnegie Mellon, wrote in his newsletter this morning. “But it is not production, and hardware is not the limit to autonomous vehicles. Software is the long pole in the tent.”

At first glance, it would seem as if the event did the trick. There were plenty of Tesla fans who were thoroughly impressed by what they saw last night and ready to declare that it was “game over” for every other player in the field. The Robovan wowed many with its Art Deco styling. And positive vibes extended to the company’s most bullish investors, some of whom participated in Musk’s theme park experience and came away forever altered.

Wedbush analyst Dan Ives, who was in attendance, dismissed any stock decline in the aftermath of the event — Tesla was trading down by nearly 9 points in early trading Friday — as a “knee-jerk reaction” that would eventually correct itself. “We strongly disagree with the notion that last night was a disappointment,” he wrote Friday, “as we would argue the opposite seeing Cybercab with our own eyes and the massive improvements in Optimus which we interacted with throughout the evening.”

It appears to be lost on some people how much things have changed since 2016, when Musk first promised that Full Self-Driving was a mere “two years away.” Many seem to be stuck in that outdated mindset that autonomous driving was an easy problem to solve and that fully driverless cars were on the cusp of taking over the world.

Since then, interest rates have skyrocketed, the buckets of ample venture capital funding have dried up, and most of the major players working on this technology have since reconfigured their timelines to account for how long it will take for self-driving cars to prove they can be safer than humans. Even Waymo, which is far and away the leader in the space, is taking things real slowly, one city at a time. It can’t promise the world; the company is still trying to figure out highways.

Musk is promising the opposite. He said that Tesla plans to launch fully autonomous driving in Texas and California next year, with the Cybercab entering production by 2026. Tesla Model 3 and Model Y vehicles with “unsupervised” Full Self-Driving would come first. But he promised that people could even buy the Cybercab for a price “less than $30,000.” Potential owners would be like shepherds, tending their flock of little driverless cabs, roaming the streetscape.

Last night, his pitch veered utopian, as images of parking lots transformed into verdant gardens displayed on the giant screens above him. (I call this “reverse Joni Mitchell-ing.”)

“We want to have a fun, exciting future,” he said, “that if you could look in a crystal ball and see that future, you’d be like, ‘Yes, I wish I could be there now.’”

It was kind of nice to hear Mr. “Dark MAGA” articulate a brighter vision for the future, but after the event, it’s even less clear how we’ll get there. We got no details about how he will overcome the enormous obstacles in his path. Here’s a quick rundown of some of the issues that went unresolved:

• Regulatory approval. Tesla will need to obtain a permit from the California DMV to operate fully driverless vehicles on public roads. And in order to do that, it will need to demonstrate that its vehicles can operate safely — which so far, the company has not done. And to produce a completely steering wheel-less Cybercab will require waivers from the federal government. That is a multi-month process, and success is far from guaranteed.
• Liability. What happens when a driverless Tesla crashes? Who takes legal responsibility? So far, Tesla has actively avoided accepting liability for its driver-assist crashes. And Musk has said he would continue to dodge liability unless there was something “endemic with the design” of the vehicle.
• Remote assistance. What happens when a driverless Tesla gets stuck somewhere? Or becomes disabled? Other AV operators, like Waymo and Cruise, have procedures in place for remote operators to try to move the vehicle out of the way. And if all else fails, they dispatch crews of technicians who come out and manually drive the vehicle. How do you do that with a vehicle without a steering wheel and pedals?
• Fleet maintenance. Tesla briefly showed an image of a snake-like robot vacuum cleaning some crumbs off the backseat of the Cybercab. But fleet maintenance is a lot more involved. Who’s going to clean the cameras during the winter or charge the vehicle when the battery runs low?
• Emergency detection. Other robotaxi companies have struggled to react to emergency vehicles, unexpected detours, and other edge cases that could arise. Tesla was investigated by the federal government for over a dozen incidents in which its vehicles using Autopilot collided with stationary emergency vehicles.

This is just scratching the surface. Kyle Vogt, the ex-CEO of Cruise, posted on X a pretty thorough list of his own questions for Tesla, most of which went completely unanswered. And this coming from a guy who was pushed out of his own company for screwing up the response to one of those hard-to-predict edge cases (a human driver struck a pedestrian, sending her flying into the path of one of Cruise’s robotaxis).

It seems unlikely that Musk will meet a similar fate as Vogt, despite fumbling the ball so badly. We’ve already seen the kinds of death and destruction that have resulted from his company’s aggressive push into autonomous technology. And so far, he’s been successful in avoiding those consequences.

But once the driver vanishes, along with the steering wheel and other controls, there will be no one left to blame but the guy who sold it.