> Waymo probably won’t have any luck sticking a vandalism charge on someone who puts a cone on the hood of its vehicles.
Good luck automating society if the public sentiment starts to really turn against these kinds of initiatives. Us nerds can argue about the merits of markets all day long, but we can’t expect the society we’re leaving behind to willingly participate in this new techno world order.
Automation is so brittle. It’s hard enough to design a fault-tolerant web, I can’t even imagine how we’re gonna do it in the physical world at scale with things like self-driving cars when they’re so easy to disable.
It’s like infosec, the attack surface of the system is far too big to properly guard against unless you’re willing to spend considerably and make the necessary sacrifices.
Markets are not even the thing. Nerds need to get to grips with ethics or at least social signaling.
E.g. how much of an burden is a service allowed to be on society at large? Most of the "disruption"-based startup models are based around exploiting some rule that societies had before that. If you have to trick society into using your service, consider you just might be on the wrong side of history.
Without knowing the details, when a nerd friend talks to me about a fictional self driving taxi start up, the first thing they would need to convince me of is that such a thing is even benefitial given the fact that it is space that is limited on city streets. Then I would have worries a out how that corp is socializing part of it's costs (by using public roads, by usings ways of attracting particular workers etc).
Don't get me wrong, those might all be ways to make good money, I don't dispute that. But as someone living in a city my stakes are different. I want to be able to get from A to B without being stuck in traffic with 100 empty cars, where the workers receive low pay and the road wear costs are paid by my taxes.
Sure, it is harder to find a way to earn money from which all people profit, but it is not impossible. If your corp is acting like a leech on society, don't be surprised if it is treated as such.
Playing video games I use to joke about people getting angry about cheats, I argued the game must be really important to them if they get so emotional.
A thing I keep thinking: A culture isn't weird until compared or confronted with a different one.
You got the impression that people don't like nerds building ridiculous things but nothing could be further from the truth.
Normals absolutely love these things, they define their being more than anything. They love to see nerds fool around AND FAIL. It's the same with investors (who are grouped along with get rich quick schemes) they love to see investors lose everything. Bitcoin is twice as exciting! Cant wait to see the pyramid implode.
I see a picture one time of a street in China and mistakenly thought it was my city in the Netherlands. It was just the context that gave it away. All the houses in those streets are different they have tiles with round holes in the side walk, the same trees and even people cycling wearing the usual... the usual cloths everyone else around the world is wearing.
This is why people take vacations far away, if you don't go far enough everything looks exactly the same, an endless regurgitation of the same mindless replication.
The answer is to do public transport with rollercoasters or turn all streets into half pipes, replace the main road with 2 looooong wave pools, use the AI to make self driving pony carts, busses with 10 floors.
I know, I know, people didn't want running water or a sewage system. Sadly for them it didn't fail but there are plenty of failures for them to laugh at, get them though their day at their useless job.
So much text and you didn't give me one substantial argument why I should change my mind.
Space is a concern in cities. It is one of the concerns. If you have the same roads-to-building-ratio public transport will always beat individual transport, because that is just basic geometry.
You like to frame this as progress, but being able to transport less people while using more resources and space isn't progress.
People are going to adjust. I remember the local forum in town with people freaking out about the Google streetview being a massive invasion of privacy and saying how they're going to chase those cars and destroy the cameras. Some people are going to have fun here and there, but I think we'll be collectively over it very soon.
> in the physical world at scale with things like self-driving cars when they’re so easy to disable.
Cars are easy to disable in general. You can permanently disable lots of cars on a random street before you get caught. This is temporary and harmless, so people aren't really serious about it.
> People are going to adjust. I remember the local forum in town with people freaking out about the Google streetview being a massive invasion of privacy and saying how they're going to chase those cars and destroy the cameras
Open Google Maps and try to streetview over Germany. People (or provacy laws) did not adjust, and I can’t say I’m sad about that - even though not having SV can be annoying.
I thought the underlying reason behind the laws and lack of StreetView in Germany was the general higher sensitivity to privacy, not just “laws”. Laws, after all, are created by elected people who should interpret the spirit of the population.
Of course people change over time to live with a technology, but I would not take it for granted that it _has_ to happen always, and not without reshaping the tech to make it more palatable to people’s concerns and values.
I'm reminded of jaywalking. Pedestrians had the right of way until the marketing departments of automotive companies decided to turn walking on the street into a crime.
Public sentiment can be easily manipulated with sufficient resources.
Social tolerance is a market. We are seeing the results of running a debt in that market.
Small scale automation is brittle. Large scale automation is far easier, especially if you're willing to purpose build infrastructure for it. This would be light rail, trains, and busses.
Also like infosec... if these cars "behaved better" then these faults would remain obscure and possibly unpatched for longer periods of time. It should make everyone wonder what other fault modes these vehicles have.
These cars that have millions of dollars in R&D sunk into them can be rendered useless by simply covering the sensors. It reminds me of the Keurig situation where they tried limiting the functionality of the machine to Keurig-signed (barcoded) K-cups. People immediately figured out how to bypass this by taping a keurig lid to the sensor so it would run every time. People are going to look for ways to get around this.
> especially if you're willing to purpose build infrastructure for it.
These companies are gonna have a hard time pulling that off politically. Public transit serves everyone (which this technology also kills), why should governments support infra built for private companies like this? The technology is not public by design. It would be political suicide.
Yes, any change that's expected to go on a social scale needs buyin from the average people. If these AVs were more expected to be used to help the more vulnerable people and in general improve everyone's life it would be cast in a different light.
Right now it has more of a technobro and rich people vibe, hoping it doesn't fall into the same kind of rabbit hole as the Google Glass and other useful innovations that just had horrendous PR and the worst people to evangelise it.
I agree with you regarding the brittleness, the small things that count.
This is the typical automation cycle. Automation is a pipeline and you keep adding/removing tasks in it.
The cone placement discovery reminds me of the feared birds, that can literally bring down an airplane when sucked into the turbine. Large steel construct vs small bird - bird wins. Large steel construct vs plastic thing - plastic wins. ;)
I’ve wondered by watching the SF rollout of more autonomous vehicles if there is central monitoring of many cars by humans.
While not fully automated, one or two people could monitor a fleet of hundreds of cars for problems with the vehicles and send one or two human agents in the field to investigate and fix.
Still stands to be wildly more efficient than a normal car ownership.
Good luck automating society if the public sentiment starts to really turn against these kinds of initiatives. Us nerds can argue about the merits of markets all day long, but we can’t expect the society we’re leaving behind to willingly participate in this new techno world order.
Automation is so brittle. It’s hard enough to design a fault-tolerant web, I can’t even imagine how we’re gonna do it in the physical world at scale with things like self-driving cars when they’re so easy to disable.
It’s like infosec, the attack surface of the system is far too big to properly guard against unless you’re willing to spend considerably and make the necessary sacrifices.