Autonomous vehicles create new problems and solve none

19 points by GrigoriyMikh ↗ HN
What is the main motivation for creating autonomous vehicles? Take delivery drones for example, what problems do they trying to solve? If we look at it from money perspective -- delivery jobs are among lowest paying. And that's the segment that all those autonomous drones/trucks are trying to take now. So basically billions of dollars going into research to solve non-existing problem.

More concerning is what comes after adoption of this technology. And this adoption almost surely will come, either due to dumping strategies, some government restrictions("Roads are Not for People" act) or, most likely, both. As a consequences, ability to find a job will loosen for majority of people. Of course, liberals(that will become only more popular, broken folk loves "free money to all" policies) will be happy to write all those people some money checks out of squeezing middle class pockets. Small businesses will not win either, since maintenance costs will be hardly much lower when summed up. The only winners -- are elites. Billionaire investors and bribed politicians.

I would be all for robots-for-business regulations. Probably, something similar to hiring folks from abroad. E.g. -- it's allowed to use robotic system for performing specific function(from the regulated whitelist of functions), only if no human applied for it in X months with salary =< X min. salaries.

Otherwise, future looks gloomy to me. Any opposite opinions?

44 comments

[ 3.3 ms ] story [ 91.8 ms ] thread
I’m not particular versed on the topic but I would have thought the primary motivators are a) increased safety (in the case of cars and transport) b) freeing up countless lost hours commuting for more productive work
There does not seem to be much appetite for freeing up hours lost commuting. We just had a massive experiment in WFH which appeared to be a massive success, yet there seems little appetite to build on it.
What do you mean by that? Is there actual data? In my region most companies that did not even consider providing WFH for everyone now treats it as their best practice. Companies I know lower their floor space permanently by around 30%.
Just what I see in London - business areas now look as full of people as they ever were.
As for safety, it would be much better to move as much of private traffic as possible to public transport. IMO, that's very liberal thing to do, and it's strange to me, why current US gov. doesn't try to improve public transportation system.

Regarding b) -- it's unclear how much of current(and potential) workforce employed in this kind of work can successfully transition into other fields of work.

> As for safety, it would be much better to move as much of private traffic as possible to public transport.

Not much better, probably slightly better than autonomous cars because the number of vehicles is smaller. But most people prefer to sit in their own cabin. Autonomous cars give you the best of both worlds. You can decide who you spend your time with or hear music as loud as you want, but are also free to do other things than driving. It's cleaner (or at least your own dirt), and you don't have to carry anything around, as you can store your stuff in the car.

> But most people prefer to sit in their own cabin

Public transport can come in very different forms. If there's demand, PT operators can build private compartments, seats with increased comfort, etc.

If you think this through, with all the comfort people usually want, you end up with individual autonomous cars. When ignoring external constraints (energy use, traffic, parking space) it’s the optimal form of transport.

Practically nobody would say something like “I’d like to share a cabin with strangers” or “I prefer to walk to a bus stop and wait there for the next vehicle” or “I like to carry my stuff around instead of just leaving it in the vehicle”.

Public transit is politically toxic in the US, in part because the middle and upper classes view it as "handouts to the poor" and don't want to use transit services with poor/homeless people on them, because "they smell" or are loud or what have you (to be fair, public transit can get pretty loud and smelly). These are problems that can be solved with sufficient funding and infrastructure, but instead US public transit services cut service coverage, remove benches from stops, raise ticket prices, and fail to do necessary maintenance.

Plus in most areas NIMBYs also block expansion. When I was in my teens attending community college back home, it was regularly discussed how there was work ongoing to expand our local public transit rail so it would reach the campus. At the time it was 3 hours on the bus (w/transfers) to reach campus so I had to get a ride or drive myself (~20 mins) to get there. ~10 years later the expansion still hadn't happened because people were fighting hard to block it.

> freeing up countless lost hours commuting for more productive work

This is the bit that worries me the most. If autonomous driving makes longer commutes comfortable then people will move out further and further. Even with EVs this will be a lot of extra energy used for transportation.

https://en.wikipedia.org/wiki/Jevons_paradox

Not really, autonomy is probably not the way to go.

One could easily drive all trains remotely, like kids playing with a toy probably since electric trains exist.

We still prefer human drivers all over the world, except for some systems. Those need to be very small or closed to be safe enough. Having a driver and personal has many advantages especially for unforeseen events. Autonomous vehicles may have the advantages in safety when nothing unforseen happens, but they can cause catastrophic chain reaktions. As critical infrastructure autonomous driving vehicles are an easy target, play them the right radar sequence and they will stop moving or endanger others.

That would be the perfect scenario. Technology that improves safety and keeps jobs.
We could even have planes flying fully autonomous. And that is airliners. Still we don't for relatively rare cases.

Actually I'm most doubtful about ships being autonomous. As they have big and complex machinery inside, with higher maintenance burden than any other type of transport. You can check plane after each flight and risk scenarios for cars and trains are not that critical. Send an other unit to pick broken one down. But ships spend weeks on seas, and everyone should work automatically? Seems bit dubious

Well, we used to have driven elevators - true they ran on tracks, but the elevator operators lobbied and harassed and warned of the doom of man......we have push button elevators. No longer is the elevator 'broken' if no driver is at work. The drivers went on to other things - we have elevators on demand. Now autonomous vehicles are somewhat similar. Driverless cars are useless, cars - as a class - spend 80-90% of their time parked. This is a massive waste of a resource. Autonomous cars, perhaps limited to well delineated paved and marked roads, can give great utility, save on time/attention. We can extend this to shared vehicles as well, if we overcome the reluctance of people to share because their trunk is full of stuff they are too lazy to deal with.
> Driverless cars are useless, cars - as a class - spend 80-90% of their time parked

You're talking about passenger cars? I'm agree that they're quite inefficient. But we don't need autonomous vehicles to solve that inefficiency. We already have perfect solution -- public transport.

There's a tipping point in population density where current modes of public transportation are not the appropriate solution. Buses, trains, subways and even cabs barely work in most suburbs and don't work at all outside of that.

Autonomous vehicles would provide a mechanism to extend the range of these services by reducing the costs of dwell time between rides and deadheading for pickups.

Yes, but does dwell time really matter? The EOL of a car is usually determined by the number of miles driven, not how long it stands without moving. On the other hand, if the (autonomous) car is not waiting for you, it needs to drive to you, driving even more miles (and wasting your time).
For private ownership no, I'm talking about its potential use for public transportation.
Uh, I live in Switzerland where buses and trains work in the entire country and in most rides - yes even country side - offer time comparable to driving by yourself. So instead of claiming something doesn't work at all, the US people could just look outside their own borders. I totally understand the political climate is very different between countries, but the technical solutions do work.
Switzerland has 8 times the GDP per square km as the US. Six times the population density and 1/90th the total road network. Switzerland has ~72,000km of roads, my local school district puts 4,500km on their bus fleet every day just picking up students.

I agree that there are no laws of physics preventing it from occurring in the US.

> Buses, trains, subways and even cabs barely work in most suburbs and don't work at all outside of that

It looks to me like mentality problem, not technology one. IMO, changing mentality of people(to refuse from excessive lifystyle/showing off) would be more beneficial for society in long term.

> Take delivery drones for example, what problems do they trying to solve? If we look at it from money perspective -- delivery jobs are among lowest paying. And that's the segment that all those autonomous drones/trucks are trying to take now. So basically billions of dollars going into research to solve non-existing problem.

Drone allow deliveries to be much faster and more "parallel". In the future, instead of having 100 delivery people, you can have 10000 drones. Drones don't tire. One drone can probably do a 100 delivery runs in the time it takes a human to do 5.

Do we really have delivery "throughput" problem? Any data confirming that?

And if there's a problem, why we need high tech solution for that? Bring more people/increase birth rate.

Yeah, at least in the US when we see stories about delivery services failing to keep up, it's usually services that have been cutting pay, increasing hours, etc and are understaffed as a result.
We certainly do here in Australia. I ordered an item from a store less than 10km from my house and it's taken over two weeks to arrive.
What do you mean by a throughput problem?

I suppose before cars were invented, you would argue that cars are useless because we already have horses and they have been getting us from A to B just fine for centuries, and there's no "data" to prove that there's any "throughput" or "speed" problem.

Maybe things take 5 days to deliver now. Maybe that's not a problem per se, but imagine if things arrive within an hour of ordering them.

> Bring more people/increase birth rate.

I'm all for increasing the birth rate, but not for the purpose of creating more "workers" to slave their time and physical labor away for "the economy".

Lots of automation has happened and is happening in the lowest paying jobs. Agriculture is the most obvious.

Another one in my city : trash collectors. We have underground collecting containers for we have a RFID 'key' to deposit in.

Once in a while, after the sensor has called home to acknowledge it is 'full', one guy shows up with a specialized truck, lifts the container from the ground and empties it into the truck. And pooff : the jobs are gone.

Self checkout is another one.

The lowest paying jobs are also the easiest to automate. And even when employees are cheap, they are still expensive and a nuisance.

imho the only way it would be beneficial is something like:

- personal cars are banned (at least from cities)

- infrastructures are completely redesigned for autonomous vehicles

- governments create fleets of public transportation autonomous vehicles you can summon at any point (from 1 seat vehicle to bigger vans)

I don't believe in the cohabitation between autonomous and human driven cars on the existing infrastructure

> personal cars are banned (at least from cities)

That's already done in some European countries. IMO, it's right direction. I'm not for banning personal transport entirely, but i think that maintenance of public roads should be done mostly from transport taxes(so tax increase is good).

HGV driving for long periods demands machine-like levels of concentration and wages for truck drivers in a post-Brexit economic environment have risen to over £50,000. I wouldn't call that a non-existent problem - of course the costs get passed on from the merchants and then to the consumer.

The concept of new technology eroding the job market isn't new either. Unemployment levels in the UK are so significantly low currently, you would not seriously suggest that the introduction of factories or computers had permanently eliminated the requirement for human labour on a massive scale. Why would autonomous vehicles pose a more serious threat?

Re: delivery drones, energy efficiency comes to mind. Now when I order lunch with delivery service, there's 1-2 tons of metal and plastic on wheels moving around with dino oil power, piloted by 80 kg of meat, to bring me a 0.5 kg delivery.

Even if the delivery vehicle ran on wheels on the roads, I'm sure it could be much lighter and energy efficient if it didn't need safely carry the driver around.

But there's usually a lot of other orders in that delivery vehicle. I assume that if drone would have similar storage capacity it would consume even more electricity.
If you take your post and replace "autonomous vehicles" with "household appliances" (Dishwashers, washing machines, tumble dryers, ...) it might be enlightening as this is something we all already got used to.

I'd say most of the "problems" you point at are not related at all to "autonomous vehicles", but to automation as a whole.

Economic disruption by automation does exist, but mostly due to our insistence of tying in the "right to live" with "participation in the economy".

There are real problems with autonomous vehicles. Primarily the amplification of existing negative externalizes of physical transport (Space, noise pollution, energy, waste, ...). Autonomous vehicles will cause an explosion in the (already insane) consumption of transportation.

> If you take your post and replace "autonomous vehicles" with "household appliances" (Dishwashers, washing machines, tumble dryers, ...) it might be enlightening as this is something we all already got used to.

I can't argue with that in numbers. But intuitively it's very different situations. In the past, only very rich households had dedicated personnel for such kind of tasks, so i guess it was not that big % of people doing that.

It is true that especially before WWII the household services market was not yet fully financialized, but that does not mean those "jobs" were not done.

The impact of that automation was hard to overestimate, as all of a sudden there were nearly twice the number of workers available, As the economic system adjusted it was the end of being able to live on one full-time income for the average family. Those hours gained from household automation where swiftly recouped in mandatory wage earning for most.

> but that does not mean those "jobs" were not done

Still, i'm not sure if any significant % of population was occupied in those jobs.

> As the economic system adjusted it was the end of being able to live on one full-time income for the average family.

By automation you mean industrialization of society? I maybe wrong, but industrialization only lead to urbanization, which separated older and younger generations which lead to "atomic family", still patriarchal in it's essence. Decay of which(as well as one full-time income per family) came only later in 20th century.

> In the past, only very rich households had dedicated personnel for such kind of tasks, so i guess it was not that big % of people doing that.

There were not many drivers either. But independent of drivers, shouldn't you also argue against dishwashers and washing machines? These are jobs that people could do manually. Low-wage jobs, so the cost of dishwashing and laundry shouldn't really matter, does it? Full-time drivers suffer serious health consequences, driving is a pretty bad job. I would assume dish washing is healthier. If we can confirm that, maybe we should automate driving and require manually dish-washing by professional dishwashers instead?

That looks like reductio ad absurdum argument. Returning professional dishwashers even not an option at this point due to depth of technology integration. Meanwhile autonomous driving is not yet here.

Anyway, as i see it, current situation is different to other similar historic ones. Maybe i'm wrong. But previously we had an abundance of low-skill jobs. Now all low-skill jobs are targets for automation.

I think that solutions like UBI are very detrimental to middle class. Since we see that higher class are very flexible in terms of tax-evasion. And i don't see how it can last for more than one generation of welfare people.

I think the most significant use of vehicles I've seen so far is: https://flyzipline.com/

Not so much replacing existing solutions but providing one that couldn't have existed before. At least as far as things go right now.

Outside of that autonomous drones have made things like mapping and other aerial sensing problems quite a bit cheaper.

My personal take is that delivery people will be amongst the last jobs to be automated as there is a lot of complexity involved. Stuff like drone delivery seems to have issues with scaling and rather than reducing the number of employees moving them from delivery jobs to a fleet of technicians to keep the drones going.

From my perspective the thing happening here is companies off-loading work onto recipients. For example quite a few of the delivery companies here now take all parcels to drop boxes and expect the customer to go pick it up. I'm not particularly opposed to that as it's probably better all round. Similarly self-checkouts seem like an improvement despite labor shifting from paid employees to the customer.

You should read up on ludities or Amish people. Ludities didn't like textile machinery. Amish like approximately 19-centry technology, no electricity etc.

In Europe we have severe shortages of track drivers. UK currently has random food shortages and fuel shortages, because EU drivers left the UK.

The fact is that people do not want to sleep on parking lots, have no shower access, irregular and long working hours etc. etc. Being a professional driver sucks as a way to spend your life.

There is a lot more enjoyable things to do.

> Ludities didn't like textile machinery.

This is a bit of an oversimplification of their position. They cared less about the machinery and more about the labor involved in using it: https://www.smithsonianmag.com/history/what-the-luddites-rea...

Indeed. Sounds very similar to the OP.

> E.g. -- it's allowed to use robotic system for performing specific function(from the regulated whitelist of functions), only if no human applied for it in X months with salary =< X min. salaries."

> They confined their attacks to manufacturers who used machines in what they called “a fraudulent and deceitful manner” to get around standard labor practices. “They just wanted machines that made high-quality goods,” says Binfield, “and they wanted these machines to be run by workers who had gone through an apprenticeship and got paid decent wages. Those were their only concerns.

I read somewhere that truck drivers are an aging population. I think Covid has resulted in some finding themselves at home and just thinking life is too short for that kind of lifestyle.
They're an aging population, yes. The reason so many have vanished is they retired. Why? They were afraid of an economic slowdown brought on by the Covid lockdown would end up with them getting laid off and losing their retirement benefits. So they took the money and ran. Now some of the retirees are being lured back to work as independent contractors. That way they've locked in their retirement benefits and still get to earn wages.