Who Watches the Waymos? I Do [video]

"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."

https://news.ycombinator.com/newsguidelines.html

Perhaps we need a euphemism for UBI: Let's call it "level-1 rich".

I once toured the one in Hamburg, which is highly recommended if you ever get the chance.

Nothing else has quite given me that feeling of being tiny next to this giant city-scale robot since.

It's not like humans are somehow put on this earth only to do certain jobs or the same job forever. Contrary to 100+ plus years of predictions, humans will never become obsolete.

https://commons.wikimedia.org/wiki/File:Line_scan_photo_of_S... (If you're confused by that page, then so was I. It contains the full photo but it's so long that it gets compressed into a line a few pixels high.)

Regular ICE cars: - air pollution from tire wear and brake dust - air pollution from exhaust - embodied carbon - noise - endangering other road users - traffic congestion - land use (sprawl) - long term health impacts (encouraging sedentary lifestyle)

Switch to all-electric and you lose a bit of noise and all the tailpipe emissions, but gain whatever emissions generated the power (sometimes solar, great, but sometimes lignite, boo), whatever environmental damage resulted from the battery materials, and probably marginal worsening of safety as the same ranger requires a heavier vehicle

Switch to self-driving and you may increase safety (feels like Waymo basically yes, Tesla probably no based on their track record with stat manipulation), but also vastly increase use, worsening congestion and land-use. The others stay the same.

So I don't understand why you're saying the externality ratio is good. From my perspective self-driving cars don't really move the needle.

Why would insurance companies lobby for that? 90% reduction in accidents means 90% reduction in premiums, which means 90% reduction in profits.

Do insurance companies have a history of lobbying for safety regulations?

I ride a motorcycle. Unless you make this illegal, then no, "manual driving" is not over anytime soon. Some people actually _like_ driving, although, none of them appear to frequent hacker news.

> there is nothing that can stop this?

We learned the lesson of over automation in aviation very quickly. The best solution is very often, turn off autopilot and autothrottles, then hand fly the plane. You don't want pilots with their heads down in a computer making changes when they can, and should, just fly the plane. It also tends to reduce over all competence and harms safety particularly during equipment failures.

> and you see already today that self-driving produces already much less accidents

The utility of these statistics are harmed by the fact that "self-driving" does not occur in all the circumstances that "manual driving" does today. They're barely past the foundation stage, in a limited operational scope (cabs), and in a limited operational area (where it never snows).

> Lets come back in 10 years! :-)

I think this technology will eventually be prevalent. I think hacker news constantly gets the time lines wrong. You're going to need 25, minimum, and possibly more like 50.

This is like saying skydiving is safer than driving because you're looking exclusively at accidents and ignoring the total number of events. There are way more cars than there are skydivers. Now in the realm of autonomous vehicles: there are way more manually driven vehicles than there are autonomous ones, so of course there's going to be significantly more accidents from manually driven vehicles.

Hint: Normies outside the tech world are not hell bent on making themselves unemployed. That phenomenon only exists among software engineers.

Insurance companies don't really care that much if accidents are more or less likely, what they care about is that they can be underwritten effectively.

The most interesting shift here as it pertains to insurance is the fact that losses might be correlated in brand new ways. Today there's no simple way for 100,000 cars to all crash a few minutes apart, worldwide, for the same reason. In the future that may in fact be plausible.

The big game changer is that the computer doesn't need sleep!!

Human truck drivers in the US are only legally allowed to drive 10-12 hours per day. Which usually means the truck is parked half the time. With a computer driving, trucks can run non stop 24/7.

This is revolutionary change! Twice the delivery speed at half the cost. Should improve a lot of things across the economy. Unless some luddites find the power to stop it.

> How will this affect transportation jobs?

The human long distance truck driver will fade away as a career. It will probably take a decade or two, so people will have time to adjust.

The transportation industry should end up substantially larger, which normally creates a bunch of new jobs/professions, but I won't pretend to know how that will play out.

Waymos are 100% reliable. If you book it and it says it's coming at X time, it will definitely actually show up at X time. No more of, driver cancelled at the last minute because they don't actually want to drive to destination Y but Uber etc gave it to them anyways and they get dinged if they just cancel instead of claiming not to be able to find the rider etc. Or driver got lost or stopped for food or gas or something so is late.

It also gets to the destination exactly when it says it will. No weird routes because of the driver's whim or driving too fast or too slow. And no chance of bad music, loud conversation in some foreign language, annoying commentary, etc.

And I want them to be profitable to run too, so they have plenty of incentive to expand the program.