Elon Musk has been very explicit in promising a robotaxi launch in Austin in June with unsupervised full self-driving (FSD). We'll give him some leeway on the timing and say this counts as a YES if it happens by the end of August.
So far Tesla seems to be testing this with employees and with supervised FSD and doubling down on the public Austin launch.
FAQ
1. Does it have to be a public launch?
Yes, but we won't quibble about waitlists. As long as even 10 [non-handpicked] members of the public have used the service by the end of August, that's a YES. Also if there's a waitlist, anyone has to be able to get on it and there has to be intent to scale up. In other words, Tesla robotaxis have to be actually becoming a thing, with summer 2025 as when it started.
If it's invite-only and Tesla is hand-picking people, that's not a public launch. If it's viral-style invites with exponential growth from the start, that's likely to be within the spirit of a public launch.
PS: I wrote the above before we learned that hand-picked invitees is indeed how this has launched so far. I meant "10 members of the public" to refer to the waitlist. I think this was clear enough at the time; not sure if it reads differently in retrospect.
A potential litmus test is whether serious journalists and Tesla haters end up able to try the service.
2. What if there's a human backup driver in the driver's seat?
This importantly does not count. That's supervised FSD.
3. But what if the backup driver never actually intervenes?
Compare to Waymo, which goes millions of miles between [injury-causing] incidents. If there's a backup driver we're going to presume that it's because interventions are still needed, even if rarely. But see FAQ 7 for a gray area here.
4. What if it's only available for certain fixed routes?
That would resolve NO. It has to be available on unrestricted public roads [restrictions like no highways is ok] and you have to be able to choose an arbitrary destination. I.e., it has to count as a taxi service.
5. What if it's only available in a certain neighborhood?
This we'll allow. It just has to be a big enough neighborhood that it makes sense to use a taxi. Basically anything that isn't a drastic restriction of the environment.
6. What if they drop the robotaxi part but roll out unsupervised FSD to Tesla owners?
This is unlikely but if this were level 4+ autonomy where you could send your car by itself to pick up a friend, we'd call that a YES per the spirit of the question.
7. What about level 3 autonomy?
Level 3 means you don't have to actively supervise the driving (like you can read a book in the driver's seat) as long as you're available to immediately take over when the car beeps at you. We'll discuss in the comments how to handle this case but I'm leaning NO because another take on the spirit of the question is whether Tesla will catch up to Waymo, technologically if not in scale at first.
8. What about tele-operation?
The short answer is that that's not level 4 autonomy so that would resolve NO for this market. This is a common misconception about Waymo's phone-a-human feature. It's not remotely (ha) like a human with a VR headset steering and braking. If that ever happened it would count as a disengagement and have to be reported. See Waymo's blog post with examples and screencaps of the cars needing remote assistance.
To get technical about the boundary between a remote human giving guidance to the car vs remotely operating it, grep "remote assistance" in Waymo's advice letter filed with the California Public Utilities Commission last month. Excerpt:
The Waymo AV [autonomous vehicle] sometimes reaches out to Waymo Remote Assistance for additional information to contextualize its environment. The Waymo Remote Assistance team supports the Waymo AV with information and suggestions [...] Assistance is designed to be provided quickly - in a mater of seconds - to help get the Waymo AV on its way with minimal delay. For a majority of requests that the Waymo AV makes during everyday driving, the Waymo AV is able to proceed driving autonomously on its own. In very limited circumstances such as to facilitate movement of the AV out of a freeway lane onto an adjacent shoulder, if possible, our Event Response agents are able to remotely move the Waymo AV under strict parameters, including at a very low speed over a very short distance.
Tentatively, Tesla needs to meet the bar for autonomy that Waymo has set. But if there are edge cases where Tesla is close enough in spirit, we can debate that in the comments.
9. What about human safety monitors in the passenger seat?
Oh geez, it's like Elon Musk is trolling us to maximize the ambiguity of these market resolutions. Tentatively (we'll keep discussing in the comments) my verdict on this question depends on whether the human safety monitor has to be eyes-on-the-road the whole time with their finger on a kill switch or emergency brake. If so, I believe that's still level 2 autonomy.
See also FAQ3 for why this matters even if a kill switch is never actually used. We need not only no actual disengagements but no counterfactual disengagements. Like imagine that these robotaxis would totally mow down a kid who ran into the road. That would mean a safety monitor with an emergency brake is necessary, even if no kids happen to jump in front of any robotaxis before this market closes. Waymo, per the definition of level 4 autonomy, does not have that kind of supervised self-driving.
10. Will we ultimately trust Tesla if it reports it's genuinely level 4?
I want to avoid this since I don't think Tesla has exactly earned our trust on this. I believe the truth will come out if we wait long enough, so that's what I'll be inclined to do. If the truth seems impossible for us to ascertain, we can consider resolve-to-PROB.
Ask more clarifying questions! I'll be super transparent about my thinking and will make sure the resolution is fair if I have a conflict of interest due to my position in this market.
[Ignore any auto-generated clarifications below this line. I'll add to the FAQ as needed.]
1,000
3.00
There's some excellent discussion happening in the comments of my new "turkla" post about this. Including a couple new predictions of my own that I'll repeat here:
Long-shot prediction: Tesla will pause their robotaxi service on September 1, citing burdensome legislation that takes effect in Texas on that day.
More confident prediction: Tesla will either not be in compliance with the new Texas law by September 1st or will comply by being officially classified as supervised level 2 autonomy.
I think either of those would yield a NO for this market but let's keep discussing and clarifying! So insanely many edge cases and gray areas here.
Bayesian-update: @MarkosGiannopoulos has downgraded my confidence in those predictions. I was jumping to some conclusions about the new law, oops. Huge thanks to Markos for doing a ton of relevant research and improving our understanding of the underlying prediction here.
Repeating again my latest from the turkla comments, here's a review of possible outcomes:
Tesla is cheating and gets caught (NO in this market)
Tesla is cheating and this gradually becomes clear as they fail to scale up as promised (probably a NO or mostly NO in this market if we can hold off on resolving that long?)
Tesla has pulled this off and the supervision is just abundance of caution (YES)
Tesla is faking-it-till-making-it but does end up making it and there's no proof they were ever faking it
Possibility 4 could be brutal for deciding a fair resolution here. Let's keep discussing it!
https://wiki.unece.org/download/attachments/128418539/SAE%20J3016_202104.pdf?api=v2
Level 3 is defined "with the expectation that the DDT fallback-ready user is receptive to ADS-issued requests to intervene" which is impossible for the safety monitor to do because they are not in the drivers seat. They can tell the car to emergency break but the robotaxi can't request them to take over.
So they are clearly above level 3 already, even with a safety monitor. But do they reach level 4 or is it somewhere inbetween, more like 3.5?
Level 4 definition:
> "The user does not need to supervise a Level 4 ADS feature or be receptive to a request to intervene while the ADS is engaged. A Level 4 ADS is capable of automatically performing DDT fallback, as well as achieving a minimal risk condition if a user does not resume performance of the DDT. This automated DDT fallback and minimal risk condition achievement capability is the primary difference between Level 4 and Level 3 ADS features. This means that an in-vehicle user of an engaged Level 4 ADS feature is a passenger who need not respond to DDT performance-relevant system failures."
The primary distinction is that at level 4 the user doesn't need to respond/take over. Like I said above that has to be true because the monitor sits in the passenger seat.
The only debatable point is that the user "does not need to supervise" a level 4 system. My argument is that since it's not necessary for operation and just an extra safety precaution (as proven by the autonomous delivery without a safety driver), this is also fulfilled.
Edit: The document also states that "Level Assignments are Nominal, Rather than Ordinal, and are Never Fractional" and "Levels are Mutually Exclusive", which makes clear that if it's above level 3 that means it's level 4. There is no such thing as level 3.5 like I suggested above.
@Toastbroti This is a hugely helpful comment. Great to get specific about the SAE standards! But using your reasoning that the autonomy levels are all-or-nothing, we could say it's "clearly" level 2 because it doesn't meet the level 3 criterion of letting the human take their eyes off the road. (To be clear, none of this is at all clear!) Here's a review of the autonomy levels:
Level 0 = a totally normal old-school car
Level 1 = old-school assistance like cruise control + lane-keeping assist
Level 2 = self-driving but a human has to have eyes on the road at all times ready to disengage the self-driving if it's about to screw up
Level 3 = the human doesn't need eyes on the road but needs to be ready to retake control immediately if the car beeps
Level 4 = the human doesn't need to be ready to either intervene or take control; the car will stop and and ask if it's confused
Level 5 = the AGI of self-driving; the car handles anything a human can
As I was saying in another thread in this market, what Tesla has at this point kind of defies those categories. It's worse than level 3 if it needs real-time supervision and better than level 3 in not needing the human to ever take over driving. I would say it's unambiguously better than level 2 and ambiguously worse than level 4. I think we're just waiting for more information to come to light about the role of the safety monitors and remote operators. Is that sounding fair?
@dreev yes that makes sense. I expect them to remove the safety monitors as they scale (they can't scale much with them), but where they draw the line is hard to guess. Even if they aren't really necessary, currently at this tiny scale they are ~free and decrease unknown risk. My best guess is that they will scale fast enough to drop them before this market resolves, but that is assuming they are going to be somewhat aggressive.
In Tesla's POV they could drop them right now or in a few months depending on how conservative they want to be. In 5 years it won't matter which month they stopped using them, so better safe than sorry. So I might lose this bet on them supervising what would be a Level 4 capable system a bit longer than I would have expected out of an abundance of caution.
I hadn't seen this Elon tweet at the time:
> "When 3:1 or more Robotaxi to Supervisor/Teleoperator ratio?"
> "As soon as we feel it is safe to do so. Probably within a month or two."
https://x.com/elonmusk/status/1939592691156914627
@Toastbroti Ah, thank you, yes, this sounds exactly right (re: your previous comment about how necessary the safety monitors actually are). I think it might be good to leave this market unresolved until we can say, with hindsight, which version was actually true. Could they have removed the supervision before end of August or not? As you can see from my own predictions, I'm NO-biased but I don't want to seize on the technicality that the safety monitors were present. We want to get to the heart of question, whether Tesla has achieved level 4 autonomy or if we're seeing basically controlled demos that wouldn't be safe without human supervision. As you say, seeing how they scale from here will help answer that question.
First autonomous delivery a day ahead of schedule. No safety monitor in the car.
https://x.com/elonmusk/status/1938682871105102254
@Toastbroti Well I'll be darned and knitted. I did not expect they were ready to do that. I'm looking wronger by the day here. What version of FSD is that?
Actually, I notice I'm confused. No new FSD version has been announced and the most recent version (13.2.9) seemed to still only be at hundreds of miles between critical disengagements. Which I took to mean that if you read a book while the Tesla was driving it would seem fine for days or weeks but eventually, in a month or something, it would kill you.
Was that wrong? Or has there been a breakthrough? (Or was Musk just like, "ok, so it crashes every 500 miles and this is a 5 mile trip, let's roll the dice!") Can the customer who took delivery of that car have it drive unsupervised?
Or maybe the first question is how sure are we that Musk is being honest in that tweet? Excerpt:
The first fully autonomous delivery of a Tesla Model Y from factory to a customer home across town, including highways, was just completed a day ahead of schedule!!
There were no people in the car at all and no remote operators in control at any point. FULLY autonomous! To the best of our knowledge, this is the first fully autonomous drive with no people in the car or remotely operating the car on a public highway.
That last sentence is pretty sus. Waymo doesn't do highways for their commercial service yet but they've been doing it with employees for over a decade (I was in the back seat of one doing so in 2011! with a human supervising in the driver's seat back then, of course). Certainly there've been plenty of Waymos on highways with no supervision for years.
I'm worried I sound like I'm grasping at straws to disbelieve this. I mean, it's definitely incredible -- either figuratively or literally.
@dreev Forehead-smack. I think I just figured it out. The car was remotely supervised with the ability to disengage and instantly transfer control to a tele-operator in case anything unsafe happened. Which, by luck, it didn't during that short trip. So, I'm now predicting, it's still level 2 and it would not be safe to override the eyes-on-the-road restriction.
@dreev That would imply that all production cars now have software for real-time teleoperation by Tesla. This would be unprecedented for a car company and for Tesla itself (no similar feature exists in the Tesla software until now).
@MarkosGiannopoulos Good point. So either this was a publicity stunt and no normal people will be getting their cars delivered this way for a while, or Tesla's had a recent breakthrough and this is for real. Maybe the following market is one to watch:
https://manifold.markets/CameronHolmes/will-tesla-offer-autonomous-vehicle
@dreev > it crashes every 500 miles
It simply doesn't. FSD 13 works fine.
https://www.youtube.com/live/B2iVZ10L3ns?si=YGBgyjx10un3aVFa&t=878
I don't know what the issue with FSD Community tracker is, but it's skewed. Maybe people disengage prematurely, maybe people who are happy with FSD just don't use it.
@MarkosGiannopoulos Wait, doesn't it sound like Musk phrased that tweet pretty carefully, only saying that remote operators never were in control during that trip? And maybe it's not that they have full tele-operation remotely but do have the emergency "stop in lane" button that they're ready to press in real time.
If so then conceivably they got away with delivering a level 2 autonomous car with no one in it. Are you game for making a market about this?
Not sure how significant this is but there's a video on ABC News that includes a clip of the safety monitor seemingly with his finger hovering over the kill switch on the touchscreen as the car navigates slowly between pedestrians. Couple key frames:
The passenger monitor sure looks like an eyes-on safety driver to me.
https://xcancel.com/teslarati/status/1937654180547821903?s=46
@WrongoPhD That one's interesting because it suggests the human supervisor does not have a physical button to trigger emergency braking. I have been assuming that the existence of a physical button is key -- that no physical controls means it counts as level 4. But mostly that's because I didn't imagine that Tesla would think it ok for the human to ever have to lunge for a button on the touchscreen.
So, I don't know, the plot thickens, I guess.
Normally level 2 autonomy means a human driver has to be ready to yank back control at any time. If a human has to supervise but their only possible intervention is a kill switch on the touchscreen, that's... I don't know, it feels like it's outside the normal categories. It's better than level 3 in some ways (the human never has to actually drive) and worse than level 3 in some ways (the human has to actively monitor at all times).
Overall I think we have to keep waiting. Something more definitive may come to light. Huge thanks for finding all these pieces to the puzzle!
My previous musing on physical kill switches, from another market:
I actually think there's a key difference in a physical kill-switch vs a button on the touchscreen. Namely, if the autonomy level is such that the car needs a human with eyes on the road at all times, ready to emergency-brake if the car's about to crash, that's supervised FSD aka level 2 autonomy. If the human has a kill switch available on the touchscreen, that's different. You better not need to look down at the touchscreen for the kill-switch button if the car's about to mow down a child that ran into the road.
In short, no one in the driver's seat plus lack of a physical kill switch (and no dead children) probably would satisfy me that these robotaxis count as level 4.
(But the example from @WrongoPhD above of a safety monitor lunging for the kill switch on the touch screen has me questioning that. Just questioning, to be clear. Nothing about any of this is obvious yet!)
@dreev You have lost the forest for the trees. The levels are supposed to indicate a self driving cars capabilities. These robotaxis are unambiguously worse than L3, in which no one needs to be paying attention to monitor the car unless the car asks for help. That's clearly not true here. The fact that the safety driver only has immediate access to a break, and possibly only clumsy access to a break, does not make the car more capable. It just makes it more reckless. The safety monitor is only in the passenger seat rather then the driver seat is for optics, not capability. The fact that we've seen a need for an intervention on only the third day of operation with only 10 cars makes that clear, especially since the passengers are super fans who might not even all choose to publish videos of mistakes.
@WrongoPhD Didn't I clarify that two comments above? But yes, if the resolution of the mystery here is that they're basically YOLOing a level 2 system, that would be pretty horrific. I guess that's what your market about a robotaxi ending up in a crash is for. And it will be frustrating for the Tesla detractors if they totally are YOLOing it but get lucky for long enough to get their shit together.
Speaking of which, I originally predicted that Tesla wouldn't pull this off until they copied all the things Musk has mocked Waymo for:
hi-def pre-mapping
the phone-a-friend feature for when the car is confused
lidar/radar
But based on what we've seen so far, I think there's a growing chance I was wrong. The only thing I'm sure of is that it's not obvious. Not yet.
@MarkosGiannopoulos Good question. It's reported in this article: https://electrek.co/2025/06/25/whoopsie-uh-oh-oh-my-heres-all-the-gaffes-and-goofs-by-tesla-robotaxi-so-far/
@dreev I would not consider Electrek a good source of information for such things. Testing in a specific area is not the same as "hi-def mapping".
@MarkosGiannopoulos I think you're right. And even they don't say "hi-def":
it’s geofenced to somewhere around 30 square miles in South Austin which Tesla spent additional time mapping and testing in