> The insight driving the program, Naga said, is that the limiting factor for AV development is no longer the underlying technology. “The bottleneck is data,” he said. “[Companies like Waymo] need to go around and collect the data, collect different scenarios. You may be able to say: in San Francisco, ‘At this school intersection, I want some data at this time of day so I can train my models.’ The problem for all these companies is access to that data, because they don’t have the capital to deploy the cars and go collect all this information.”
You can’t be the CTO of Uber wanting to do AVs, and get the data collection requirement shockingly wrong.
Waymo is able to deploy with less (but targeted and high quality) data collection by having world class simulation capabilities. Not that they haven't collected huge amounts of data as it's no doubt important (I've heard their onboard storage is transferred and emptied every few days), it's just not a bottleneck. They have the most efficient operation in the AV industry.
The best example of why data collection isn’t the bottleneck is Tesla. They boast about billions of miles of data, yet they’re struggling to put out fully autonomous vehicles.
> When they want data about a school intersection in SF at a certain time of day, they just... synthetically generate it and simulate
I think it's more about detecting changes to the world. You need boots on the ground, so to speak, to see that new speed limit sign or the new lane paint. The Waymo vehicle can no doubt react to changes in the world when it encounters them, relaying them back to the mothership, but it's better to know about them in advance.
Most AVs, definitely Waymo vehicles, are self mapping. They can detect environment changes and relay it to the entire fleet. That's because they map using the same vehicles as the fleet.
Yeah I'm not so sure this CTO is on the mark here, but to be fair, I do think some of this IRL long tail/edge case data is important for Waymo. The simulation software is super interesting to me - the real world can be so chaotic, and even if they could generate every possible real life case, there needs to be validation on whether the Waymo driver is responding in the optimal way. They certainly haven't solved this problem, you can see some of their growing pains in all of these articles - floods in Austin, more and more interactions with emergency vehicles that first responders seem to believe are getting worse, etc.
Tesla on the other hand has billions of miles of data, yet because there is a limit to camera-only techniques, that data isn't that useful is it? They have no ground truth data to evaluate their camera system on, which is why sometimes you see those Teslas driving around with lidar rigs mounted on them. Going camera-only is just asking for trouble.
I agree real world data is important for Waymo. I didn't mean to say it wasn't, so I've edited my comment to reflect that. It's just that data is not some magic bullet to achieve self driving like Tesla and others suggest.
Of course, Waymo still has much more room for improvement. But it's much more efficient to supplement less but higher quality IRL data with large amounts of synthetic data, than to run a million data collection vehicles 24x7 because most IRL data is boring and useless.
Waymo said 6 years ago they simulate 20 million miles every single day [1]. Clearly, it's working for them given their scale of deployment right now.
Although most of the real-world data is probably boring, collecting more of it likely makes discovering rare edge cases more likely. But since they happen rarely, I imagine that after discovering them, they would then need to figure out how to simulate them.
I find the idea of learning from simulated data so unintuitive. How can you radically improve your model with just your model? I take it people do it, so it must work, but i just don’t understand it at all.
Well there's a world simulation model and then the driving model.
You can imagine improving i.e. a specialized math model (problem in, theorem out) with a normal LLM that knows lots of problems and theorems generally.
> The best example of why data collection isn’t the bottleneck is Tesla. They boast about billions of miles of data, yet they’re struggling to put out fully autonomous vehicles.
Well, TBF, the tesla data was complete garbage with earlier vehicles. They had cheap and somewhat bad cameras in the earlier vehicles that was only somewhat recently updated. And even then, I don't think Tesla is at the end of their hardware journey. I think they don't think that either, which is why they've gone to a subscription only model for self driving vehicles.
Waymo, on the other hand, has gathered less data, but more high quality data. They do the expensive mapping of a city which is a big part of why their vehicles have early on been able to do some pretty impressive feats. The drawback is getting that high quality data takes a lot of time and resources.
Didn't they need the data from the 200 million miles or so from actual driving before they could get to the generative model though? Data isn't everything, as you point out with Telsa (mainly because they decided to forego using lidar it would seem), but it is pretty fundamental.
> Waymo might very well be missing specific kinds of data (e.g more incidents/accidents, near-collisions etc)
Accidents and near-collisions are exactly the kind of scenarios perfect for simulation. You don't test them out in the real world and risk injuries/deaths. You need to have confidence they're handled before you deploy.
Immediately after leaving this thread I saw a post on Bluesky where someone was discussing the GUARD act and used AV to mean "age verification." It's out of control!
"Our goal is not to make money out of this data" is doing a lot of heavy lifting for a company that just committed $10B to robotaxis and is taking equity in the same AV companies it would be supplying.
The actually interesting part isn't the sensor grid, that's years away and has real consent, compensation, and regulatory problems nobody's talking about. It's shadow mode: letting AV companies sim-run their models against millions of real Uber trips without putting a car on the road. That works today. That's the product.
The sensor grid is the press release. Shadow mode is the business.
Also completely absent from this article: do the drivers whose cars become "rolling data collection platforms" get anything? A cut? A notification? A commemorative badge? Uber has a rich history of finding creative ways to extract value from its driver network, so I'm sure they've thought carefully about this.
Uber was working on self driving ten years ago. They had cars on the road loaded with cameras and sensors specifically to collect data. Then they negligently killed a woman crossing a street.
This isn't a pivot, this is them trying to sheepishly reenter the race they were dramatically ejected from.
I feel like they should have done this 6 years ago. Most AV companies already have tons of their own data today. But how would it work to install expensive LIDAR sensors on privately-owned vehicles?
Exactly my take as well. This would have been the right diversification move a decade ago.
Uber did invest early in self driving back in 2015, but in 2018 there was a fatality which pretty much deleted their whole program. And looks like it's taken them way too long to try picking it back up.
FWIW, a large fraction of Uber drivers aren't actually driving their own personal cars, at least around me nowadays. They're either rented or some sort of fleet vehicle (complete with TCP #)
I'm honestly surprised that Tesla never took advantage of all the cameras in all its cars to do some kind of mapping project. I always thought that was incredibly valuable data. Sort of an automatically crowd-sourced street view.
Some people would tell me that they do, but only for training their internal self-driving AI.
I'm not sure about the privacy implications. You say "all its cars" but you actually mean "all its customers cars". The relationship between Uber and the cars/drivers is fairly different.
It’s not, but this guy is in probably 70% of threads posting what he thinks is the same submission. I’ve seen several instances where the posts are different but he just links it and says it’s a duplicate anyway. If he wants to be a mod so bad he should just apply for a job imo.
Isn't Uber also replacing itself? If you use your human drivers to train other companies' robot taxis, aren't you gonna ruin both the human driver service and the data collection service?
Shareholders won't like the plan of limiting growth. That's a really interesting point, what's the plan for when human drivers can no longer generate new data and they've already sold they what they've managed to collect to all the car markers who want to buy?
I'm not too worried about it. Yeah, it's bad that people don't understand how labor organizing works. It's bad they're not willing to stand up to shitty employers and take a little risk to make life better. But in this particular case the fear is totally illusory. It's just another silicon valley conman selling some warped "dream" that probably won't actually materialize[1]. "Autonomous Vehicles" are nowhere near production ready, and they're not going to be any time soon. Wake me up when a serious truck or car manufacturer starts rolling them out en masse, then I might start to get worried about it. Until then, it's just about the same category as flying cars--sure, we have these hexacopter contraptions which can (barely) lift a single person for 20min. Not interesting.
[1] Here's how you know:
“Our goal is not to make money out of this data,” Naga said. “We want to democratize it.”
> Wake me up when a serious truck or car manufacturer starts rolling them out en masse, then I might start to get worried about it
I think enough people haven’t been in a Waymo to realise that the technology is basically here, and that we’re like 10 to 20 years of doubling away from AVs doing tens of millions of trips a day in America. By the time anyone has invested in true mass production of AVs, we’ll already be so far down that path that the policy deck will be dealt.
My Subaru can lane keep in a Wyoming blizzard. There isn’t some unsolved technical problem with snow for any system with radar, i.e. anyone who isn’t Tesla.
Keep in mind that like a fifth of Americans and half of humans live somewhere is rarely or never snows.
I guess once they demonstrate it working smoothly and profitably in cities like Boston, NY, Detroit, etc. I'll be more concerned? Given that it doesn't really even work too well in places with good signage, lane markings, etc--not to mention no weather--I'm not worried yet. And frankly I'm done with "the future". If it doesn't deliver results now, it's not real. Until it's real, it's nothing to get alarmed about.
> Given that it doesn't really even work too well in places with good signage, lane markings, etc
Works fine in Phoenix, Miami and Los Angeles. Plenty of neighbourhoods there have non-existent, defaced and degraded signage and markings.
> Until it's real, it's nothing to get alarmed about
I don't think there is anything to be alarmed about, period. Driving is a silly job when you think about it. We made these machines to do our bidding, not enslave us behind their wheels.
My belief in a smooth roll-out is reinforced by those who would probably oppose AVs also not believing it's real. Once the first factory mass manufacturing AVs breaks ground, any limited local opposition can be preëmpted.
People don’t understand the slow motion horror movie that this is becoming. Labor demand begets population growth for all of human history. Demand conditions set the stage for population growth. Labor surplus set the stage for population declines. Again, this has been true for all of human history.
So what are we walking into? Not 8-11 billion happy cows. A crisis. People deciding not to reproduce. The human population declining. The irony as we achieve a technical pinnacle while justifying our own extinction by choice. The great filter as it turns out is actually capitalism, a race to business efficiency against all else including the incentives of your very own species. This is the mind virus.
Preface: I am personally NOT into anti-growth ideas, and I also think it’s super alarming that the West especially seems to be intent on wiping itself out by lack of having kids.
But that said, supposing we are looking at 60 years from now having a few billion fewer people on Earth, just by attrition (lack of replacement) that is not automatically bad. We could afford to shrink in population - if there’s a floor to that contraction. If indeed there are way too many people in a decade for the available human jobs, then it could be the equilibrium is just a lower population. Which could be temporary - who knows what the future could bring, such as possible space colonization, which may need more humans and also give people the hope that I think Gen Y and Z have lacked, which is one reason for their low repro rates.
Look at it from the very high level point of view of the ~500 or so richest people on the planet that rule everything: Planet Earth is basically a vacation resort for them that they control, and the remaining 7,999,999,500 of us are the staff that serve them and keep them entertained and comfortable. If that can still happen with only 6 billion, 2 billion, or 50 million of us, they are not going to care. When their wealth stops coming from other people's labor, they'll just dispose of us.
The complete destruction of the human through exploitation and control, as seen in the article, is a major reason people are too unhappy to start families.
The worst part? Most people don't even know why, so there's never a general public reaction to fix it.
this whole argument depends on the supposition that if brith rates ever drops below replacement rates, then that inexorably implies the extinction of the species. whether or not now is a good time, at some point growth has to stop. and there are plenty of conceivable social arrangement that are perfectly workable with a constant population size.
the only real argument for continued growth in to preserve the current structure of investment. that's your great filter, and it will result in economic collapse which isn't the same as extinction.
It’s here and it’s been here for decades - it’s just finally impossible to ignore or wave away
Gig workers are self-chattelizing because there is no floor to the depravity that society will accept, and an endless supply of people who will chattelize themselves for a moment of pleasure
> The majority of people have absolutely no foundational belief for their actions
The majority of people have absolutely no choice for their actions
FTFY
Especially given that for an increasing number of folks the alternative to doing what you're told is starvation (food insecurity), homelessness, and death (from lack of healthcare).
> majority of people have absolutely no choice for their actions
The poorest Americans remain among the richest people in the world and across human history. Many people don't have material choices. Almost every American does.
There's no advancement of the discussion to be had.
If I may play his game:
He writes to boost his own ego. By way of making his claim, he seeks to appear wise and wizardly, for who else but he could have made such an astute observation and present it with such confidence; it must certainly be true and he must certainly be better than us.
And who would make such a claim, but the one who is pure enough to see through the muck and see the truth of the claim? He must not be one of the egoist consuming hedonists! He must be outside of them to have seen them!
Or not, perhaps he is one of these egoists and he knows it, and will happily admit it. And by doing so, he will raise himself up even further. For he knows his faults and he is not ashamed to have us know them too. We shall soon see.
In the end, he will soon have more sex. (He already got the attention.)
It’s necessary to recognize the limitations of the individual’s and the system they comprise
If it is required to take action in the world as a person or anything that has the capability to do things then by function there is energy required to take action
If the actions of that population are detrimental to the global population, because local measurement of global externalities is ignored, then you have a population that is self-destructive simply because it cannot communicate or coordinate across all of the places that’s having impacts
The systems you currently live in are antihuman…yet entirely composed of humans acting for their own individual gain without the capability of considering the collective whole
Is it anti-human to suggest that existential threat that stems from a total failure of the biological organism to coordinate at the scale that it’s individual impacts have ?
Is it antihuman to claim that the population does not have the capability of being able to counter its own accelerating self-destructive behavior?
> an endless supply of people who will chattelize themselves for a moment of pleasure
Or perhaps they "chattelize" to survive?
There's not much pleasure to be had from gig work apart from the freedom to perhaps choose your own hours and perhaps be free of a human boss. Both of which are quite the opposite of chattelizing, in the short term.
Define survival first and we can have a conversation
They have birthday parties and loving embraces in deprived ghettos that have community solidarity
The most beautiful human interactions I’ve ever seen are in the absolute most deprived poor places including when I was working in the fucking Balad hospital in 2010
I won't debate the definition of survival with a "tired old" developer (oh sorry, "founder") who's idea of virtue is summed up in their own quote regarding creating yet another app for the Apple/Google chattel system:
"I rarely get to see my kids. That's a risk you have to take."
I’ve been waiting for someone to pull that one out as a gotcha…
But hey good for you for doing a bunch of searching about me personally (because I intentionally use my name so that people can do precisely what you’re doing) which indicates that I have triggered you to the extent where you’ve taken time out of your day to go and look me up personally
I have ADHD so me searching the internet is like breathing. Nothing special.
And your quote, inflammatory marketing slop that it is, is top and center in the images after searching your name once. "A bunch of searching" is not required. There's not much out there about you that requires digging into, just the usual founders' must-haves (crunchbase profile, paid write-ups, personal blog etc). Nothing special there either.
But please do enjoy the extra attention from me. Because that is special. To me.
It's more productive to discuss and bring to light the floor's underwriters than it is to blame gig workers for "chattel-izing themselves for a moment of pleasure".
> I asked if he had ever heard of collective bargaining or knew about unions and he said no.
collective bargaining or unions do not prevent technological progress, but merely retard it in the hopes that their members can benefit at the cost of progress for everyone else. Look at dock workers and how they tried to prevent automation with unions.
Dock workers are really the best example. We should have been automating at least a decade ago. I don’t know why folks would think unions or collective bargaining should be used to prevent automation. You will just lose on the medium to long term.
Reminds me of when dockworkers resisted the shift to cargo containers. Those ports ultimately lost business in the end.
You can’t be the CTO of Uber wanting to do AVs, and get the data collection requirement shockingly wrong.
Waymo’s bottleneck has never been data. When they want data about a school intersection in SF at a certain time of day, they just... synthetically generate it and simulate: https://waymo.com/blog/2026/02/the-waymo-world-model-a-new-f...
Waymo is able to deploy with less (but targeted and high quality) data collection by having world class simulation capabilities. Not that they haven't collected huge amounts of data as it's no doubt important (I've heard their onboard storage is transferred and emptied every few days), it's just not a bottleneck. They have the most efficient operation in the AV industry.
The best example of why data collection isn’t the bottleneck is Tesla. They boast about billions of miles of data, yet they’re struggling to put out fully autonomous vehicles.
I think it's more about detecting changes to the world. You need boots on the ground, so to speak, to see that new speed limit sign or the new lane paint. The Waymo vehicle can no doubt react to changes in the world when it encounters them, relaying them back to the mothership, but it's better to know about them in advance.
It'll shock you to know that you can simply get this from governments, some even provide this in API form
Tesla on the other hand has billions of miles of data, yet because there is a limit to camera-only techniques, that data isn't that useful is it? They have no ground truth data to evaluate their camera system on, which is why sometimes you see those Teslas driving around with lidar rigs mounted on them. Going camera-only is just asking for trouble.
Of course, Waymo still has much more room for improvement. But it's much more efficient to supplement less but higher quality IRL data with large amounts of synthetic data, than to run a million data collection vehicles 24x7 because most IRL data is boring and useless.
Waymo said 6 years ago they simulate 20 million miles every single day [1]. Clearly, it's working for them given their scale of deployment right now.
[1] https://waymo.com/blog/2020/04/off-road-but-not-offline--sim...
You can imagine improving i.e. a specialized math model (problem in, theorem out) with a normal LLM that knows lots of problems and theorems generally.
Well, TBF, the tesla data was complete garbage with earlier vehicles. They had cheap and somewhat bad cameras in the earlier vehicles that was only somewhat recently updated. And even then, I don't think Tesla is at the end of their hardware journey. I think they don't think that either, which is why they've gone to a subscription only model for self driving vehicles.
Waymo, on the other hand, has gathered less data, but more high quality data. They do the expensive mapping of a city which is a big part of why their vehicles have early on been able to do some pretty impressive feats. The drawback is getting that high quality data takes a lot of time and resources.
Is that the right kind of model for this particular application?
Also, Uber’s data might be useful for eval, not training (e.g « here is how Waymo would behave vs human drivers therefore it is safer »)
Accidents and near-collisions are exactly the kind of scenarios perfect for simulation. You don't test them out in the real world and risk injuries/deaths. You need to have confidence they're handled before you deploy.
The original title says "self-driving" and that's much more clear.
I saw the title and thought it can't be AV, they must mean AI and made a typo.
This isn't a pivot, this is them trying to sheepishly reenter the race they were dramatically ejected from.
Uber did invest early in self driving back in 2015, but in 2018 there was a fatality which pretty much deleted their whole program. And looks like it's taken them way too long to try picking it back up.
Real-world data spoils faster than a gas station banana.
If your AV company is relying on data from six years ago, you're going to kill someone.
https://www.getnexar.com
Can the drivers charge a monthly late for hosting the sensors?
Don’t they [1][2]?
[1] https://www.privacyinternational.org/examples/1929/tesla-lea...
[2] https://electrek.co/2020/10/24/tesla-collecting-insane-amoun...
I'm not sure about the privacy implications. You say "all its cars" but you actually mean "all its customers cars". The relationship between Uber and the cars/drivers is fairly different.
Seems par for the course. Nintendo turned legions of Pokemon Go players into unpaid sensor grids for delivery robots.
Uber torches 2026 AI budget on Claude Code in four months
https://news.ycombinator.com/item?id=47976415
It sounds like a terrible business plan.
He said he “didn’t care and besides what was he going to do about it anyway, it’s going to happen no matter what”
I asked if he had ever heard of collective bargaining or knew about unions and he said no.
I think we’re only about another generation before the only purpose for human labor is to train and check the outputs of a machine.
[1] Here's how you know:
I think enough people haven’t been in a Waymo to realise that the technology is basically here, and that we’re like 10 to 20 years of doubling away from AVs doing tens of millions of trips a day in America. By the time anyone has invested in true mass production of AVs, we’ll already be so far down that path that the policy deck will be dealt.
My Subaru can lane keep in a Wyoming blizzard. There isn’t some unsolved technical problem with snow for any system with radar, i.e. anyone who isn’t Tesla.
Keep in mind that like a fifth of Americans and half of humans live somewhere is rarely or never snows.
Couldn't come soon enough [1].
> Given that it doesn't really even work too well in places with good signage, lane markings, etc
Works fine in Phoenix, Miami and Los Angeles. Plenty of neighbourhoods there have non-existent, defaced and degraded signage and markings.
> Until it's real, it's nothing to get alarmed about
I don't think there is anything to be alarmed about, period. Driving is a silly job when you think about it. We made these machines to do our bidding, not enslave us behind their wheels.
My belief in a smooth roll-out is reinforced by those who would probably oppose AVs also not believing it's real. Once the first factory mass manufacturing AVs breaks ground, any limited local opposition can be preëmpted.
[1] https://waymo.com/blog/shorts/back-to-boston/
So what are we walking into? Not 8-11 billion happy cows. A crisis. People deciding not to reproduce. The human population declining. The irony as we achieve a technical pinnacle while justifying our own extinction by choice. The great filter as it turns out is actually capitalism, a race to business efficiency against all else including the incentives of your very own species. This is the mind virus.
But that said, supposing we are looking at 60 years from now having a few billion fewer people on Earth, just by attrition (lack of replacement) that is not automatically bad. We could afford to shrink in population - if there’s a floor to that contraction. If indeed there are way too many people in a decade for the available human jobs, then it could be the equilibrium is just a lower population. Which could be temporary - who knows what the future could bring, such as possible space colonization, which may need more humans and also give people the hope that I think Gen Y and Z have lacked, which is one reason for their low repro rates.
The complete destruction of the human through exploitation and control, as seen in the article, is a major reason people are too unhappy to start families.
The worst part? Most people don't even know why, so there's never a general public reaction to fix it.
the only real argument for continued growth in to preserve the current structure of investment. that's your great filter, and it will result in economic collapse which isn't the same as extinction.
It’s here and it’s been here for decades - it’s just finally impossible to ignore or wave away
Gig workers are self-chattelizing because there is no floor to the depravity that society will accept, and an endless supply of people who will chattelize themselves for a moment of pleasure
The pleasure of being an Uber driver? Wouldn’t the better analogy be survival for most gig workers?
it is simply how do I get more money sex property attention etc…
There are no monks door dashing
The majority of people have absolutely no choice for their actions
FTFY
Especially given that for an increasing number of folks the alternative to doing what you're told is starvation (food insecurity), homelessness, and death (from lack of healthcare).
The poorest Americans remain among the richest people in the world and across human history. Many people don't have material choices. Almost every American does.
Might want to remove the "Wizard" from your bio, it'd be far more accurate.
people want to get money so that they can have an “enjoyable life” very few indeed dedicate themselves to a virtue or an ethics above pleasure
“enjoyable life” as defined through pretty much the entirety of written human history consists of sex and play
So yes humanity is demonstrably composed of egoist consuming hedonists
Okay. How does this advance the discussion? Obviously then that means opposing that means being anti-human.
If I may play his game:
He writes to boost his own ego. By way of making his claim, he seeks to appear wise and wizardly, for who else but he could have made such an astute observation and present it with such confidence; it must certainly be true and he must certainly be better than us.
And who would make such a claim, but the one who is pure enough to see through the muck and see the truth of the claim? He must not be one of the egoist consuming hedonists! He must be outside of them to have seen them!
Or not, perhaps he is one of these egoists and he knows it, and will happily admit it. And by doing so, he will raise himself up even further. For he knows his faults and he is not ashamed to have us know them too. We shall soon see.
In the end, he will soon have more sex. (He already got the attention.)
If it is required to take action in the world as a person or anything that has the capability to do things then by function there is energy required to take action
If the actions of that population are detrimental to the global population, because local measurement of global externalities is ignored, then you have a population that is self-destructive simply because it cannot communicate or coordinate across all of the places that’s having impacts
The systems you currently live in are antihuman…yet entirely composed of humans acting for their own individual gain without the capability of considering the collective whole
Is it anti-human to suggest that existential threat that stems from a total failure of the biological organism to coordinate at the scale that it’s individual impacts have ?
Is it antihuman to claim that the population does not have the capability of being able to counter its own accelerating self-destructive behavior?
Or perhaps they "chattelize" to survive?
There's not much pleasure to be had from gig work apart from the freedom to perhaps choose your own hours and perhaps be free of a human boss. Both of which are quite the opposite of chattelizing, in the short term.
They have birthday parties and loving embraces in deprived ghettos that have community solidarity
The most beautiful human interactions I’ve ever seen are in the absolute most deprived poor places including when I was working in the fucking Balad hospital in 2010
Virtue does not come from work
"I rarely get to see my kids. That's a risk you have to take."
I’ve been waiting for someone to pull that one out as a gotcha…
But hey good for you for doing a bunch of searching about me personally (because I intentionally use my name so that people can do precisely what you’re doing) which indicates that I have triggered you to the extent where you’ve taken time out of your day to go and look me up personally
And your quote, inflammatory marketing slop that it is, is top and center in the images after searching your name once. "A bunch of searching" is not required. There's not much out there about you that requires digging into, just the usual founders' must-haves (crunchbase profile, paid write-ups, personal blog etc). Nothing special there either.
But please do enjoy the extra attention from me. Because that is special. To me.
collective bargaining or unions do not prevent technological progress, but merely retard it in the hopes that their members can benefit at the cost of progress for everyone else. Look at dock workers and how they tried to prevent automation with unions.
Reminds me of when dockworkers resisted the shift to cargo containers. Those ports ultimately lost business in the end.