Elon Musk is about to masterfully move the goalpost on Tesla Full Self-Driving

Elon Musk is about to move the goalpost on Tesla’s Full Self-Driving (FSD) program in a masterful way that will allow him to claim a win.

Don’t be fooled.

I asked Grok, Musk’s “truth-seeking AI”, to list all of the CEO’s timelines for Tesla achieving self-driving and then compared them to reality:

• 2015: Predicted full autonomy by 2018.
  • Reality: Not achieved; Tesla cars still required human supervision.
• 2016: Claimed full autonomy from LA to New York by 2017, all Teslas as robotaxis by 2020.
  • Reality: Did not happen; FSD remained in development.
• 2018: Full self-driving capability in “3 to 6 months”.
  • Reality: Missed, FSD still required human supervision.
• 2019: Early access FSD by year-end, full unsupervised driving in 2020.
  • Reality: Limited beta release in 2020, but not unsupervised.
• 2020: Very close to Level 5, quantum leap in FSD.
  • Reality: FSD Beta launched, but still Level 2 with human oversight.
• 2021: Full self-driving, 1 million robo-taxis by year-end.
  • Reality: Neither achieved; FSD remained in beta.
• 2022: Full self-driving by end of 2022 or May 2023.
  • Reality: Did not reach this milestone; FSD still not fully autonomous.
• 2023: Reiterated confidence in achieving full self-driving.
  • Reality: No full autonomy achieved; continued FSD improvements but still supervised.
• 2024: Announced unsupervised FSD in Texas and California for Q2 2025.
  • Reality: As of early 2025, this has not yet been launched.
• 2025: Specified launch of unsupervised FSD in Texas for June 2025.
  • Reality: As of now, this has not been confirmed to have occurred; the timeline is still within the projected future, so no definitive comparison to reality can be made yet.

That’s about as embarrassing as it gets, but many Tesla supporters still don’t care because they believe that now it’s going to finally happen.

As the last point states, Tesla is still within the latest timeline of “unsupervised FSD in Texas in June 2025.”

The problem is that what Tesla is planning to launch in Austin in June has very little to do with what Musk has been promising and selling to Tesla FSD buyers since 2016.

As the latest data shows, Tesla FSD is still far from unsupervised self-driving in customer vehicles, which was promised, but it has improved significantly in the last few months. The combination of the improvement and the fact that Musk can’t take many more losses with missed FSD timelines has pushed Tesla to find a solution: Waymo.

Musk has pooh-poohed Waymo’s approach to self-driving for years. He claimed its geo-fenced, mapped, teleoperation-supported approach wouldn’t scale.

Yet, that’s almost exactly what Tesla is about to launch in Austin this year.

The CEO confirmed that Tesla’s plan is a “paid unsupervised self-driving ride-hailing service using an internal fleet of Tesla vehicles” in Austin in June.

We reported that Tesla was looking to hire people to work in teleoperation to support its self-driving vehicles shortly after announcing its plan for unsupervised ride-hailing services in Texas and California last year.

The planned teleoperation, combined with the service being limited to Austin, points to Tesla launching a geo-fenced service where it will optimize FSD performance in Austin and use teleoperation to support the vehicles.

That’s exceptionally close to Waymo’s product, which has been available in many cities for years, including in Austin more recently.

As for the long-anticipated unsupervised self-driving capability in all customer vehicles produced since 2016, it looks like Musk is too scared to share a timeline after being consistently wrong for a decade.

Electrek’s Take

I can almost guarantee what will happen: Tesla will launch this project and claim to have achieved “unsupervised self-driving.”

Elon and his Tesla influencer simps will pump this up while blurring the line between this product and FSD in customer vehicles to give the impression that Tesla is still a leader in self-driving.

When, in fact, Tesla will only have achieved what Waymo delivered years ago.

Tesla won’t be closer to delivering what it promised and sold to owners since 2016: unsupervised self-driving capable of robotaxi driving in customer vehicles.

As of the latest data, Tesla FSD v13 is achieving about 500 miles between critical disengagement while Tesla’s own stated goal to be safer than humans is to surpass miles between collision with human drivers, which is at 700,000 miles, according to NHTSA.

This program in Austin is no more than a diversion, a moving of the goalpost, to give Tesla an impression of a win in self-driving and distract owners who have bought FSD and have been promised unsupervised self-driving capability for years.

Then you had the HW3 situation into the mix, and you have quite the mess.