ra7

It seems like the real news is Jeff and Sanjay are leaving Google, and Demis is effectively replacing Jeff as Chief Scientist for all of Alphabet.

The bigger deal is the departure of Jeff and Sanjay, rather than Demis moving into a different role.

> and thus not useful

Again, I’m not suggesting this. Bottleneck has a specific meaning. It means Waymo is limited by not having the ability to collect data. Well, clearly that’s not true because Waymo already has a reasonably scaled deployment across a dozen cities that no one else has and can handle millions of scenarios.

Real world data is absolutely required, but more of it doesn’t give you magical self driving ability as Uber’s CTO suggests. If it were the case, you’d see Tesla achieve fully autonomous driving years ago.

“Real world data is not the bottleneck” != “Real world data is not useful”

No one is suggesting the latter.

Mapping and simulation have very different purposes. Doesn’t look like you’re familiar with the basics of AV technology. Explains a lot why you’re confused about how real world and simulated data is used.

That’s literally how it works right now, so yeah.

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.

> 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.

IIRC, they had clocked 20 million real world miles before starting to scale their deployment. But they were also driving 20 million miles in the simulator every day: https://waymo.com/blog/2020/04/off-road-but-not-offline--sim...

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.

[1] https://waymo.com/blog/2020/04/off-road-but-not-offline--sim...

> 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’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.

This is fascinating. I feel like this is converging into the concept of a traditional "IDE". So much of your setup reminds me of IDEs indexing, doing static analysis, building ASTs, etc. before a developer starts writing code.

> Before Waymo deploys in a new city, it deploys a huge fleet of cars that spend months of driving completely supervised, presumably to construct a detailed LIDAR map of the city.

Not entirely true. From their recent "road trips" last year, the trend is they just deploy less than 10 cars in a city for a few weeks (3-4 weeks from what I recall) for mapping and validating. Then they come back after a few months to setup infrastructure for ride hailing (depot, charging, maintenance, etc.) and start service.

We all know Tesla likes to play smoke and mirrors game with vehicle numbers — they have 300+ "robotaxis" but only 7 of them are unsupervised [1], and they shut down when it rains [2].

So let's use a metric that unequivocally shows who is 'winning'. I'm confident Waymo will have more paid rides per week than Tesla at the start of 2027 (I'll give you 2028 if you want). No other metric indicates scale better than passenger trips. If you have more robotaxis or you are in more cities, it will show up in the trip count.

I'll give $1000 to a charity of your choosing if Tesla beats Waymo in this metric. Fully unsupervised trips only, does not include trips with a safety driver or a monitor in a passenger seat, none of the usual games they like to play.

[1] https://robotaxitracker.com/?provider=tesla

[2] https://x.com/ethanmckanna/status/2022803049551372395

Nothing says confidence like a prediction with an unspecified timeline.

Waymo drives 4 million miles every week (500k+ miles each day). Vast majority of those collisions are when Waymos were stationary (they don’t redact narrative in crash reports like Tesla does, so you know what happened). That is an incredible safety record.

> Tesla beating Waymo

Heard this for a decade now, but I’m sure this year will be different!

There’s also one where Tesla hit a parked truck:

“13781-13644 Street, Heavy truck, No injuries, Proceeding Straight (Heavy truck: parked), 4mph, contact area: left”

They don’t have remote drivers. Your own link says that.

> The Waymo Driver does not rely solely on the inputs it receives from the fleet response agent and it is in control of the vehicle at all times.

…

> The Waymo Driver evaluates the input from fleet response and independently remains in control of driving.

These videos from Waymo shows what kind of guidance they provide:

https://youtube.com/watch?v=T0WtBFEfAyo

https://youtube.com/watch?v=elpQPbJXpfY

Notice how the system itself reasons about the scene and asks for help with possible options.

This whole story is a nothingburger. The only “news” here is that the operators are in Philippines.

> Tesla is executing the strategy that most quickly scales to 100% of the population.

So, uh… where is this “scale” then? This “strategy” has been bandied about for better part of a decade. Why are they still in a tiny geofence in Austin with chase cars?

Waymo is doing it right now. Half a million rides every week, expansion to a dozen new cities. Tesla does a few hundred in a tiny area.

Scale is assessed by looking at concrete numbers, not by “strategies” that haven’t materialized for a decade.

This is 3D LiDAR output (multimodal) from 2D images.

We know Waymo reduced their LiDAR price from $75,000 to ~$7500 back in 2017 when they started designing them in-house: https://arstechnica.com/cars/2017/01/googles-waymo-invests-i...

That was 2 generations of hardware ago (4th gen Chrysler Pacificas). They are about to introduce 6th gen hardware. It's a safe bet that it's much cheaper now, given how mass produced LiDARs cost ~$200.

The novel aspect here seems to be 3D LiDAR output from 2D video using post-training. As far as I'm aware, no other video world models can do this.

IMO, access to DeepMind and Google infra is a hugely understated advantage Waymo has that no other competitor can replicate.

> Persistent directory at ~/.claude/projects/{project-path}/memory/, persists across conversations

I create a git worktree, start Claude Code in that tree, and delete after. I notice each worktree gets a memory directory in this location. So is memory fragmented and not combined for the "main" repo?

Yes, it provides external validation for the valuation. Otherwise, Alphabet can simply "self value" Waymo at a funny amount like $1T.

There's also a strategic partnership angle in these rounds. For example, Magna and Autonation were early investors in Waymo. Magna operates Waymo's factory in Arizona to upfit their vehicles with sensors, Autonation (the huge dealership/service network) is the maintenance partner.

In general, the Alphabet playbook is that projects "graduate" out of Google X, and are expected to operate as a standalone company, including being responsible for raising funds.

You’re being vague and hand-wavy yet again.

Do you stop at every double parked vehicle when you’re driving? A Waymo would never move if it always predicted pedestrians would jump in front of it out of nowhere or the car next it would swerve at 65 mph. It’s physically impossible to stop in time for many accidents, unless you’re already stopped.

What do you mean “fast enough”? You can’t predict something that doesn’t exist i.e. not visible to the sensors. A Waymo wouldn’t move at all if it assumed people would always jump out of nowhere.

Even if you detect “fast enough”, there are physical limits for braking and coming to a stop.

There have been many instances of Waymo preventing a collision by predicting pedestrians emerging from occlusion. This isn’t new information at all for them. Some accidents are simply physically impossible to prevent. I don’t know for sure if this one was one of those, but I’m fairly confident it couldn’t have been from prediction failure.

See past examples:

https://youtube.com/watch?v=hubWIuuz-e4 — first save is a child emerging from a parked car. Notice how Waymo slows down preemptively before the child starts moving.

https://www.reddit.com/r/waymo/s/ivQPuExwNW — detects foot movement from under the bus.

https://www.reddit.com/r/waymo/s/LURJ8isQJ6 — stops for dogs and children running onto the street at night.

Tesla invests $2B in xAI, SpaceX merges with xAI. It's just blatant self-dealing all the way down to artificially prop up valuations.

If you want American companies to not outsource any jobs AND have full foreign market access, get ready to get market access revoked from places like India. They’ll just incentivize their local companies to compete, and Amazon has plenty of local competition there already.

Amazon themselves have experienced in the past how heavy-handed Indian regulators can be.

It’s not a zero-sum game anymore. You cannot have only one side (US companies) capture 100% of the value.