Travis Kalanick, the former CEO of Uber who helped pioneer the ride-hailing industry, says Waymo is “obviously” ahead of Tesla in the robotaxi race, and that Elon Musk’s company needs a “ChatGPT moment” for its vision-based autonomous driving to catch up.
The comments came during an appearance on the All-In podcast this week, where Kalanick also revealed he’s jumping back into the self-driving game himself with a new robotics venture called Atoms.
Kalanick’s take on the robotaxi race
Kalanick told the All-In hosts that Waymo has pulled ahead of Tesla as the two companies roll out autonomous vehicles across the US. He noted that Waymo’s key challenges going forward are “manufacturing and scale and urgency and fierceness”, not its core technology, which he views as proven.
Tesla, on the other hand, is pursuing a fundamentally different approach. While Waymo’s vehicles rely on an expensive suite of cameras, lidar, and radar sensors, Tesla is betting everything on a vision-only system using the cameras already built into its EVs. Kalanick characterized this approach as chasing a “ChatGPT moment”, a sudden breakthrough that would make cameras-only autonomous driving work at scale.
The distinction matters. Waymo’s 6th-generation hardware cut its sensor count by 42% and brought per-unit hardware costs below $20,000 — a 50% reduction from the previous generation. The company now operates fully driverless rides in 10 US cities, delivers around 400,000 paid rides per week, and is targeting 1 million weekly rides by the end of 2026.
Tesla’s position is far more modest. The company operates roughly 35 vehicles in Austin with in-car safety supervisors and has managed to run only a single truly unsupervised robotaxi in a limited section of the city. Tesla has reported 15 crash incidents to NHTSA since the Austin program launched in June 2025, and the company redacts crash narrative details, making independent safety analysis impossible.
Kalanick is getting back in the game
The former Uber CEO isn’t just commentating — he’s building a competitor. Kalanick recently unveiled Atoms, a robotics company that emerged from eight years of stealth operations under his real estate venture City Storage Systems, which owns the ghost-kitchen operator CloudKitchens.
Atoms is targeting the food, mining, and transport industries. Kalanick is also on the verge of acquiring Pronto, the autonomous vehicle startup focused on industrial and mining sites created by his former Uber colleague Anthony Levandowski, the same engineer who was convicted over autonomous driving trade secrets stolen from Google’s self-driving program (he later received a presidential pardon).
Reports from The Information indicate that Uber is backing Kalanick’s new self-driving effort, and Kalanick has told people close to the venture that he intends to pursue a more aggressive rollout of autonomous technology than Waymo.
That’s a bold claim given Waymo’s current trajectory. The Alphabet-owned company raised a massive $16 billion round at a $126 billion valuation in February and expanded to four new US cities last month alone. The company delivered 15 million rides in 2025 and reports 90% fewer serious injury-causing crashes than human drivers.
Electrek’s Take
I agree with most of Kalanick’s comments, but I strongly disagree on Waymo’s key challenges being “manufacturing and scale and urgency and fierceness.” In fact, I feel like Waymo’s biggest challenge is to continue on its path without being dragged into a race with Tesla and others that would put scaling in front of safety.
The fact that Kalanick says that at the same time that he claims he wants to catch up to Waymo is scary.
To some degree, Waymo is still limited by safety. Yes, it could likely operate in additional US markets with its current capabilities, but I think it still needs more testing and validation for the next stage of real-world growth, including longer-distance trips and more highway driving.
There’s a risk of falling into Tesla’s game and moving too fast. I think Waymo is doing well and needs to keep moving in that direction.
The “ChatGPT moment” framing is generous to Tesla, honestly. It implies that a breakthrough is plausible and would change the game overnight. That might be true, vision-only autonomous driving at scale would be transformative. But Tesla has been promising this breakthrough for years. Musk first promised “Full Self-Driving” capability by 2018, then robotaxis by 2020, then a million robotaxis on the road by 2020, none of it materialized on schedule.
To me, it looks like Tesla needs many more years and/or resources to cover as many tail-end scenarios as possible, or just give up and start using radar and lidar.
FTC: We use income earning auto affiliate links. More.
Comments