NOTE: This market includes any autonomous ride in a Tesla, not just the robotaxi service!
Waymo claims 700k fully autonomous rides in 2023, which was more than ~0 Tesla rides that were fully autonomous in 2023.
To define fully autonomous, I'll use the following definition:
No human operates any direct controls inside the vehicle like the steering wheel or pedals
No human in the car actively monitors the driving being ready to intervene
Must be of a non-trivial distance (e.g. crossing the parking lot does not count)
Can't be in a fixed track, like the Las Vegas Loop
See also: /JamesGrugett/will-tesla-serve-more-fully-autonom
I will delay resolution until both companies report results or until the result is extremely clear, or until it's January 2028, where I'll make an informed guess.
1,000People are also trading
@TeslaBatteries @teslacoil68k @Teslatheshiba @neotesla @realelonmusk @elonmusk @MuskElonn @elonmusk99Rl @Elon @ElonsCalculator They're doubting you
Update: Thanks @WrongoPhD! May the best win. I will try my best to judge it fairly. I expect the resolution will be obvious either way.
@JamesGrugett The only risk I feel is whether you'll be an honest arbiter for this bet. That said, you certainly bought yourself credibility by taking your lumps in the 2025 market. I think the gulf between going 10k miles without intervention and going 100k miles is much larger than you realize, and even 100k miles wouldn't be enough for me to bet my life on. Time will tell.
Waymo claims 700k fully autonomous rides in 2023, which was more than ~0 Tesla rides that were fully autonomous in 2023.
How would Tesla’s 2025 efforts count towards this market? IIUC, the summer launch in Austin is likely to be considered insufficient autonomy for @dreev ’s market, but he says Tesla’s latest efforts could pass his threshold. Would the summer robo taxi rides have counted as fully autonomous for the purposes of this market? What about e.g. the guy who reports FSD from coast-to-coast?
(There’s some ambiguity in resolution but I think a fair amount of that ambiguity already applies)
@Ziddletwix That's a good question. I would count those rides for the most part. Maybe a small percent had the technician actually intervene. I know @dreev has a complicated thing where just the ability for them to possibly intervene means that it's not autonomous, but the same is true of Waymo with their remote operators. The clearer rule is: did anyone actually intervene in the autonomous ride?
I'll state it again for clarity: I expect millions of these autonomous rides to be Tesla owners using their private cars not in the robotaxi fleet. So the guy who went coast to coast counts.
I'd also expect a large chunk of these private rides to be nearly autonomous, in that they used the steering wheel like once to leave the driveway or whatever. Those don't count for this market. I recognize that it's going to kind of hard to figure out this stat, but we can do our best to estimate it haha.
@JamesGrugett I don't think we need to quibble about this, but I don't think the coast to coast ride should count based on your criteria "No human in the car actively monitors the driving being ready to intervene." The coast to coast ride definitely had someone monitoring the car ready to intervene, regardless of whether an intervention was required. In fact, FSD disables if you're caught not paying attention.
@JamesGrugett Doh, sorry for making it complicated! The reason I've emphasized "ability to disengage" instead of the cleaner "actual disengagements" is just due to insufficient mileage. Fatalities with human drivers happen something like once every 100 million miles. Waymo is beyond that number with no fatalities so far. When Tesla gets there we won't need to quibble about actual vs counterfactual engagements.
It's all about 9's of reliability. If someone drives coast to coast and never has to disengage the self-driving, that might just be that nothing crazy happened to occur on that journey like a kid running out in the road.
I also need to clear up confusion about tele-operation. Waymo has a thing called remote assistance but it's night-and-day different from tele-operation. Waymo is level 4, not level 5, and I think that's the right standard to use for Tesla as well. (Briefly, level 4 allows for the car getting confused and autonomously stopping and asking for help, as long as the human is never in control in real time.)
But this is another thing we can stop worrying about at scale. With the small-scale Austin robotaxi launch, we're suspicious that Tesla might be cheating with tele-operation. When it scales to millions of cars (especially privately-owned ones) it'll be safe to assume Tesla is not using tele-operators, if they ever were.
So I think the confusion and seeming contradictions are all specific to trying to decide if today's robotaxis count as autonomous. I felt a little vindicated on this when Musk said in the November shareholders meeting that unsupervised self-driving was very close, maybe a month away. I'm not sure how to read that other than as an admission that the robotaxis in Austin were, as of November, still depending on the passenger-seat safety monitors having a finger on a kill switch and/or tele-operation. And, again, I do think it's possible for the human monitors to have been necessary even if there never happened to be any interventions so far. At scale, lack of actual interventions can absolutely be taken to imply autonomous driving.
@Simon74fe tesla has way better unit economics, waymo is growing pretty slowly in fleet size compared to what tesla could manage if they had the software and regulatory parity, and so the market is in my mind asking if they will get that before EOY? like invoking the god of straight lines on graphs seems particularly mistaken here, and the graph is misleading by inviting ppl to do this
@Simon74fe Nice! I hope you'll keep updating that throughout the year. @Bayesian is quite right that the reason the probability for this market is as high as it is is that Tesla may soon crack actual level 4 autonomy and they're poised to surge past Waymo whenever that happens. But Waymo's growth, slow as it is, is exponential in the literal sense. So this could all be pretty exciting and I think watching that graph adds to the excitement.
Like you said, if there's anything misleading about it besides omission of Tesla's potential to go hyper-hyper-exponential when they have vision-only level 4 autonomy, that would be good to hear. (Cards on table: I'm a Tesla skeptic myself and made a killing in the 2025 version of this market but am far less confident for this market.)
@Bayesian I don’t think the graph misleads anyone. It’s just very funny.