Hacker News new | ask | show | jobs
by morpheos137 1258 days ago
With all due respect the hubris (and folly) of Silicon Valley pursuing driveless cars and trucks as well as things like uber is astounding.

It will fail in the market because the technology does not exist and will not exist in the foreseeable future unless or until something like AGI is perfected.

Driving is not a closed domain problem. There will always be fat tail circumstances where statistical methods will break down.

Furthermore driving, freight haul and taxicabs are NOT natural monopolies. They are not areas where technology can give enduring market power that society will tolerate (except perhaps in the case of regulatory / institutional capture). Instead they are low margin, established, service- commodity businesses.

The point of trucking is to safely and efficiently get goods from A to B. Until machines / automation can do it cheaper than humans it will not be viable.

There is no shortage of cheap low skilled labor in the USA.

Why invest in hugely expensive automation in an open problem domain that you can not even perfectly model when you can just pay a driver $25 an hour to do it better?

There is so much more easily attainable low hanging hanging fruit.

Why not automate warehouses?

Trains? Waitstaff? Insurance agents? Real estate agents?

Triage nursing?

All these are probably more attainable.

The driverless delusion reeks of sophomoric understanding of the problem domain fueled by credulous media and group think.

Personally I wouldn't hire anybody for a thinking role who believes that driverless trucks or taxis are viable marketable product for the open road. It is a good heuristic to separate critical thinkers from bandwagon riders.

It is literally an easier technical problem to land a man on the moon than to design an autonomous road vehicle that can do better and be more cost effective than humans in all circumstances.

Hopefully we will stop talking about this foolishness in a couple years.

https://blog.piekniewski.info/2021/12/18/farcical-self-delus...

3 comments

> Driving is not a closed domain problem. There will always be fat tail circumstances where statistical methods will break down.

Some of it is. Walmart has already deployed some self-driving semis in Arkansas that go between a warehouse and a store. There are a significant number of routes like this, always taking the same path back and forth between two places in a relatively controlled environment.

> Why invest in hugely expensive automation in an open problem domain that you can not even perfectly model when you can just pay a driver $25 an hour to do it better?

Driving time is a big one. There are strict limits on how many hours drivers can be driving on a given day. That means you've got a huge, expensive machine sitting idle for most of the day. Self-driving can get you double the usage of an expensive asset.

> Why not automate warehouses?

This is happening already. https://www.wsj.com/story/amazon-takes-steps-toward-warehous...

> It is literally an easier technical problem to land a man on the moon than to design an autonomous road vehicle that can do better and be more cost effective than humans in all circumstance.

Nobody's saying it has to work in all circumstances. You can choose where to deploy self-driving vehicles. If you stick to the American Southwest, you still have an enormous opportunity, but you eliminate one of the big challenges for autonomous vehicles - snow and other difficult weather.

> Driving is not a closed domain problem. There will always be fat tail circumstances where statistical methods will break down.

It doesn’t have to solve all cases to be immensely useful. All it needs is to be able to recognize when it’s unable to safely have control with enough advance notice to delegate to a human. Remotely piloted aircraft have been a thing for a long time; seems like we could do the same with trucks, so that humans only need to do the driving a small % of the time. I also suspect that current AI will likely respond to emergency situations better than humans do on average, because they don’t get tired and have superior reaction time.

You can't just dump control to the person sleeping or playing marvel snap in the driver seat as your out. With aircraft there are attentive people on standby and more than one. With remote aircraft people are way less likely to die if something goes wrong. You also start really lowering the training level of humans.

I'll alter one of my favorite jokes to explain.

I want to go out like my Tesla ai did, in a bit of confusion and reanimated from rom, not screaming no no no no like the people in the car.

Tho truly with a lot of these the people will go out in their sleep.

Importantly with aircraft the time you have to correct an error that requires human pilot intervention before you crash is far greater than in a car. This is basic critical thinking... But so many people say we have autopilot why not driverless cars? Well obviously driverless cars are orders of magnitude more sensitive to error conditions on open roads than a plan flying in empty space.
I wasn't comparing autopilot in an aircraft to autopilot in a car. I was comparing remote piloting of an aircraft to remote operation of a vehicle.
the paren'ts exact same point holds, how long to swap an attentive person in. Have you done support chat? it will immediately be one person to 55 cars, and add in the latency, they're not stopping that tesla move left and hit the breaks issue let alone a 30 car pileup.
I said "All it needs is to be able to recognize when it’s unable to safely have control *with enough advance notice to delegate to a human*"

IMO, there are road conditions where this is possible. For black swan cases like random wind gusts, tire blowouts, boulders falling onto the highway etc. I think AI would be better on average *currently* (again, just clarifying what I already said).

I'm not saying that you attempt FSD with the copout that a driver needs to stay attentive and be ready to intervene. I'm saying that there is a high proportion of truck driving that's completely mindless, and it's possible to have a system that recognizes the difference.
Driverless trucking encompasses a large numbers of usage in between full autonomy everywhere and no autonomy at all.

For example, fully autonomous trucks which only work on highway with special equipment on the side could already be worth it. Long haul drivers can’t drive without stopping, can make mistakes and it’s not a very pleasant job. If you could only use a driver for the last few miles, that’s a huge gain.

You are mistaking a failure of your imagination for an absence of possibilities.