Technology

That's what happens when you have a reasonable sensor suite with LIDAR, instead of trying to rely entirely on cameras like Tesla does.

Why are we still doing this? Just fucking invest in mass transit like metro, buses and metrobuses. Jesus

Also, Note that this is based on waymo's own assumptions, that's like believing a 5070 gives you 4090 performance...

This would be more impressive if Waymos were fully self-driving. They aren't. They depend on remote "navigators" to make many of their most critical decisions. Those "navigators" may or may not be directly controlling the car, but things do not work without them.

When we have automated cars that do not actually rely on human being we will have something to talk about.

It's also worth noting that the human "navigators" are almost always poorly paid workers in third-world countries. The system will only scale if there are enough desperate poor people. Otherwise it quickly become too expensive.

Because they are driving under near ideal conditions, in areas that are completely mapped out, and guided away from roadworks and avoiding "confusing" crosses, and other traffic situations like unmarked roads, that humans deal with routinely without problem.
And in a situation they can't handle, they just stop and call and wait for a human driver to get them going again, disregarding if they are blocking traffic.

I'm not blaming Waymo for doing it as safe as they can, that's great IMO.
But don̈́t make it sound like they drive better than humans yet. There is still some ways to go.

What's really obnoxious is that Elon Musk claimed this would be 100% ready by 2017. Full self driving, across America, day and night, safer than a human. I have zero expectation that Tesla RoboTaxi will arrive this summer as promised.

They're super conservative. I rode just once in one. There was a parked ambulance down a side street about 30 feet with it's lights one while paramedics helped someone. The car wouldn't drive forward through the intersection. It just detected the lights and froze. I had to get out and walk. If we all drove that conservatively we'd also have less accidents and congest the city to undrivability.

I am once again begging journalists to be more critical ~~of tech companies~~.

But as this happens, it’s crucial to keep the denominator in mind. Since 2020, Waymo has reported roughly 60 crashes serious enough to trigger an airbag or cause an injury. But those crashes occurred over more than 50 million miles of driverless operations. If you randomly selected 50 million miles of human driving—that’s roughly 70 lifetimes behind the wheel—you would likely see far more serious crashes than Waymo has experienced to date.

[...] Waymo knows exactly how many times its vehicles have crashed. What’s tricky is figuring out the appropriate human baseline, since human drivers don’t necessarily report every crash. Waymo has tried to address this by estimating human crash rates in its two biggest markets—Phoenix and San Francisco. Waymo’s analysis focused on the 44 million miles Waymo had driven in these cities through December, ignoring its smaller operations in Los Angeles and Austin.

This is the wrong comparison. These are taxis, which means they're driving taxi miles. They should be compared to taxis, not normal people who drive almost exclusively during their commutes (which is probably the most dangerous time to drive since it's precisely when they're all driving).

We also need to know how often Waymo intervenes in the supposedly autonomous operations. The latest we have from this, which was leaked a while back, is that Cruise (different company) cars are actually less autonomous than taxis, and require >1 employee per car.

edit: The leaked data on human interventions was from Cruise, not Waymo. I'm open to self-driving cars being safer than humans, but I don't believe a fucking word from tech companies until there's been an independent audit with full access to their facilities and data. So long as we rely on Waymo's own publishing without knowing how the sausage is made, they can spin their data however they want.

edit2: Updated to say that ournalists should be more critical in general, not just about tech companies.

We always knew good quality self-driving tech would vastly outperform human skill. It's nice to see some decent metrics!

But when it does crash, will Google accept the liability?

No shit. The bar is low. Humans suck at driving. People love to throw FUD at automated driving, and it's far from perfect, but the more we delay adoption the more lives are lost. Anti-automation on the roads is up there with anti-vaccine mentality in my mind. Fear and the incorrect assumption that "I'm not the problem, I'm a really good driver," mentality will inevitably delay automation unnecessarily for years.

"After 6 miles, Teslas crash a lot more than human drivers."

Unprofessional human drivers (yes, even you) are unbelievably bad at driving, it's only a matter of time, but call me when you can do it without just moving labor done by decently paid locals to labor done remotely in the third world.

"Waymo reports that Waymo cars are the best"

I had a friend that worked for them in the past. They really aren't that impressive. They get stuck constantly. While the tech down the line might be revolutionary for people who cannot drive for whatever reason right now it still needs a LOT of work.

How are they with parking lots, tho'?

What’s tricky is figuring out the appropriate human baseline, since human drivers don’t necessarily report every crash.

Also, I think it's worth discussing whether to include in the baseline certain driver assistance technologies, like automated braking, blind spot warnings, other warnings/visualizations of surrounding objects, cars, bikes, or pedestrians, etc. Throw in other things like traction control, antilock brakes, etc.

There are ways to make human driving safer without fully automating the driving, so it may not be appropriate to compare fully automated driving with fully manual driving. Hybrid approaches might be safer today, but we don't have the data to actually analyze that, as far as I can tell.