California's Governor Vetoes State Ban on Driverless Trucks
wired.com
wired.com
We would obviously need to collect lots of data with safety drivers in place to determine how often they are necessary and how many accidents (both lethal and non-lethal) the trucks get into to determine if they really are dangerous, but this is a totally reasonable take from the head of the CFL.
> "We will continue to fight to make sure that robots do not replace human drivers and that technology is not used to destroy good jobs.”
Using union power to protect jobs just for the sake of protecting them is the thing that labor unions do that most people don't support. See the recent writer's strike for a good example -- many supported their desire for better wages and working conditions, but feelings were much more mixed on support for allowing AI to replace writers.
And depending on the source you look at, Truck Driver isn't even in the top 25 for men (it's mostly construction, HVAC, Pipe fitting, and related jobs).
And in fact one source I looked at said that software engineering employs more men than truck driving (but that source was dubious).
https://www.census.gov/library/stories/2019/06/america-keeps...
> Automated trucking means possibly lower prices, and AI entertainment does to.
We currently have inflation driven across a wide swathe of goods, driven primarily by corporate greed, and streaming services jacking up prices while putting out endless reboot and IP farm garbage. I’m not sure what would ever lead you to this conclusion.
For lowering prices vs. eating the profits, I think a bit of both happens in most industries. There are plenty of examples where better tech constantly leads to lower prices - basically most PC hardware like SSDs today. IMO these kinds of things never get media attention but yeah prices do get lower in many areas. The dynamics of whether they extract profits or pass on savings are complex and vary over time depending on competition etc. but over the long term its hard to argue that some industry will get more efficient but no one else will start competing with the same tech to bring prices down. For self driving cars we have 2-3 players in the field too which is a good sign.
Overall this argument has been rehashed for different technologies many times. Horse drivers vs. cars, scribes vs. typewriters blah blah. Its difficult to argue we would have been better off today by protecting jobs over letting the technology help everyone eventually. Personally i think this transition will be rough though, a lot of people including programmers are going to lose their jobs and we might have years of pain ahead before things stabilize. But in the long term yeah its better than just hiding AI in a box and pretending like nothing happened.
My problem is that people are increasingly blaming technology for the failures of the government. Self driving trucks SHOULD make things cheaper. Having a robot do things that people used to do SHOULD help us all in the long run the same way washing machines do.
But when it doesn't, people have such low expectations of the government that they don't even consider that option. Like are we supposed to just stop all scientific and technological progress that could endanger jobs now, just freeze society in the state its at now?
The way I see it we need both levers, science and government to make progress. Lately the sentiment is we should burn down the half that works because the other half doesn't work. I think that's the wrong approach.
[1] https://ised-isde.canada.ca/site/competition-bureau-canada/e...
I am shocked how little sympathy people have for e,g, teachers, who impact their kids every day, and have been consistently crapped on for decades on this continent. General public's overall sympathy to unions and/or truckers is limited.
Unfortunately, most people's attitude is "I don't have that, so they shouldn't either!" or "My life is hard, theirs should be too!!!". It would be more productive and compassionate to say "I don't have what they're asking for either, maybe I should do something about that and we'll all get better", but sadly that's not how it tends to work :-<
https://katv.com/news/nation-world/chicago-teachers-union-le...
As parent, you basically have to find tutors, enroll kids into AoPS/Russian Math, etc… I, for one, stopped understanding what is it that these teachers actually do?
Teachers have definitely been "crapped on" for a long time, but they fervently argue for the system from which the crap flows.
When her job became more overhead than actually teaching, that's when she called it quits. Now she volunteers most of the day at school trying to help the teachers in any way she can to make their jobs easier.
if the windows of the truck are tinted and you can't tell whether or not an AI is driving and if the script isn't signed with an author's signature and you can't tell who wrote it then who how will you even know to be upset or not?
Radical idea here: How about some places accept autonomous trucks, and some places accept human-operated ones, and then customers can choose which places they frequent? If they truly would side with truck drivers, they should go to stores that only used human-operated trucks without the need for legislation, right?
Certainly there are good reasons why SFFD is not keen on Waymo and Cruise.
Self-driving (and AI in general) has a tendency to go catastrophically wrong for impossible to determine reasons (imagine if each hallucination was a potential traffic accident), in a way that is hard to predict and is essentially open ended in terms of consequence. Humans do very stupid things much much less often than fairly stupid things - no such correlation is offered by AI.
I don't think that's true. I live in SF and am not aware of a single catastrophe owing to self-driving cars, let alone enough such catastrophes to form a statistical sample from which we could reliably infer tendencies.
A Verge article [0] has some discussion on this with regard to a motorcyclist that follows this pattern, and FortNine also covers this in a characteristically well-made video [1].
There's also the problem of autonomous (electric) cars being too quiet when they first begin moving, but that's easy to solve.
So I'm not entirely sure about their safety being "proven", but that's largely because the problem domain is enormous and the data somewhat open to interpretation. Overall "proved" seems too certain a word as of now, without narrowly defining the scope.
[0] https://www.theverge.com/2022/7/27/23280461/tesla-autopilot-...
Not to discount the obvious potential safety issues, just noting that Tesla is taking a very different approach that is currently pretty far from being able to complete trips consistently, let alone be measured up against human drivers.
It could be slightly lower but its still pretty high. This poll says 74% of likely voters support a prohibition on AI replacing humans for writing TV shows compared to 67% support for the strike overall.
It is a lot lower than the 87% support for "appropriate compensation" for reruns and streaming, but I think that is a much less contentious question than the AI writing prohibition one. By definition practically everyone agrees that they deserve "appropriate" compensation, the debate is over what is appropriate.
https://www.dataforprogress.org/blog/2023/8/17/a-majority-of...
Now, I'm not thrilled with the idea of ending human jobs without giving those people a way to survive. Even if I weren't sympathetic to those hard-working people who are ready and willing to do the tough jobs that keep society running -- and I hope it's obvious that I am -- enlightened self interest means that I don't want all of to be unemployed and hungry. That's bad for everyone. I also wish we shipped more freight via train, which is cheaper and way more environmentally friendly. Making it easier and cheaper to ship even more via truck is probably the wrong path to optimize.
But all that aside, I still think this is a well intended but ultimately wrong solution. Frankly, it seems like it'd be cheaper and more efficient to pay those drivers to stay home rather than pay them to perch in a self-driving truck.
Eventually, you simply have to go to work for a roof over your head, food and healthcare.
It doesn't really matter if you like it or not. If you current job ceases to exist you will get another job, or you will die.
Just be upfront: removing jobs is drowning people. That's it. Don't comfort yourself with "retraining" programs like they mean anything. They don't. Acting like you're giving them an oar is just insulting them on top of taking their livelihoods, so I understand the anger.
I would argue that any long-haul trucker or coal miner will have a much higher quality of life when those jobs go away and they re-train for something where they don't have to be away from their families doing dangerous things.
No, you don't. Plenty of retired people don't go to work and yet they have all of those things.
of course, retirement is an option.
But we're talking about people that need to re-train, so I'm assuming here we're talking about people who are not yet ready to retire
You make it sound like there are only two choices:
1. Confronting the politics of building a social saftey net in America.
2. Retarding innovation, wasting human potential, etc. etc.
Yet there is a perfectly good third solution that has worked countless times in history when technology has made certain jobs redundant, and will without a doubt continue to work long into the future.
3. Re-skill and get a different job.
> Eventually, you simply have to go to work for a roof over your head, food and healthcare.
I'm glad we finally agree that there are at least three choices, that building a safety net is one of them, and that going to work so that you have a roof, food, and healthcare isn't the only one.
We’re not there with truck driving yet, but to me it seems inevitable. I wouldn’t consider it as a brand new career worth starting today if my intent were to retire from it.
But there's "is" vs "ought". It ought to be the case that these hard-working people can earn a decent living doing the jobs they spent many years in. It is the case that a lot of those jobs are disappearing as the natural result of technological and societal changes. My grandpa couldn't have delivered ice in downtown St. Louis with a horse and cart today if he wanted to. That job no longer exists. In the case of coal mining, most people don't want those jobs to exist anymore: each ton of coal pulled out of the ground is nearly 3 tons of CO2 gas into the air (purely from the burning process, not counting the effort to mine it).
I don't have the answers here. I don't know what we should be doing. I sure don't want my BIL to lose his career. But what is the kind, empathetic thing to do? We've gone from zero to having self-driving cars zipping around San Francisco in a few short years. What will they be like in 10 years? 20? I don't know when it'll happen, but it seems absolutely inevitable that at some point in the near future, autonomous vehicles will be better drivers than we are. After that, do we build a little passenger cabin onto automated vehicles so that we can pay "drivers" to pretend to operate them? Even if we did, how long can we keep that going? I don't know. I just can't see a path where that's the new state of affairs forever.
Sometimes love means having hard, uncomfortable conversations. I think that's where we're at today with jobs like mining, and where we'll be soon with other careers.
I don't understand what you mean. Can you clarify?
> These features can drive the vehicle under limited conditions and will not operate unless all required conditions are met.
If one of these conditions is locality then the vehicle cannot be driven autonomously outside that locale. If a freight truck needs to go from City A to City B -- as they often do -- but the autonomous software only works in City A, then the truck won't be driven autonomously even if it's otherwise capable of autonomous driving.
So I slightly misspoke by saying it's "not autonomous" but going into level 4 vs. level 5, a freight truck is not functionally autonomous until it's level 5. It can be driven autonomously in some conditions but such conditions are almost certainly not all conditions that human freight drivers encounter daily.
(FWIW, I also entered the conversation late and am mostly responding to this: "I'm not persuaded autonomous trucks need the level 5 defined in the way the link you provided defines it.")
I'm just thinking, what if those conditions don't prohibit a truck from operating outside of a locale, but rather, they prohibit a truck from operating on anything but certain prescribed routes, on interstates, connecting ports and rail terminals and distribution centers, in the right lane only, in daylight, and in good weather conditions only? Is that level 5? It doesn't sound like level 5 to me. It sounds like level 4. And yet, it also sounds like it would still be very useful to someone running a fleet of trucks.
So again, I'm not persuaded level 5 is necessary for autonomous trucks. Some kinds of trucking would be excluded until we get to level 5, but that still leaves some useful trucking that could be handled by level 4.
>.> ohio
There's a reason why we don't have people remote-controlling trucks today.
The writing is on the wall that self-driving vehicles are coming (taxis, transport trucks, etc.). We could argue all week about the timeline for when exactly, but the important thing is they are coming.
The really important part of this is how will society handle the "loss of jobs". Are we already preparing for this eventuality? Are we gearing up to re-train truck drivers, are we gearing up to not have thousands of taxi drivers per city and all the stuff that comes along with that (added taxes, medallions, etc.)
Do we want the protection of jobs to limit our use of technology?
Should we block progress just so people can keep jobs that are not healthy for them in the first place?
Laws can be revised.
They/we kill 10s of thousands of people on US roads each year.
Instead of fighting for jobs that are becoming obsolete solely so those people don't lose jobs, maybe their union should be working on answers to this question. I'm not saying I support these people losing jobs but, like you said, should "It'd take away jobs" be a reason we should never progress technologically? Even if it stands to potentially vastly decrease the amount of accidents non-truckers die in? While they die because they're forced to overwork and overwork means sitting for most of their waking hours?
Instead of spending money on these kinds of battles the union could be working towards figuring out programs that could set these truckers up for other jobs or towards some sort of strategy that will cover their expenses until they get on their feet.
But, then again, most of those would mean their priority is their union members and not just making money. Because one way they will likely end up spending any dues money on their members with the end game eventually being dissolving the union when there are no more truckers. aka, no more money.
These trucks are gonna be used for depot to depot transfers with human drivers taking over for urban / last mile destinations.
How about, before you approve a bill for human-less autonomous driving, you first use the software with human oversight until the software is old.
Edit: after a quick Google, Embark Trucks[1] claims that they are the first to have driven 100k miles. That's a factor of 10 less than what Waymo claims to have driven. And there will be smaller outfits that will want to cut corners to get a piece of the pie.
Anyways, quite a difference. I’m not sure where the threshold should be set for proving safety. I recall the number for provable safety being something like 20bn+ miles, so no one is close if it’s set there.
I don't really understand what you want here. If you claim that a person in one of these vehicles will increase safety I am sceptical. Concentration falls off drastically when you're not actively driving.
Ships and planes are different as most of the times margins of reaction time is not seconds, but minutes... Giving enough time to person overseeing to re-orient themselves.
For example I've driven between Los Angeles and the Bay Area late at night, and between central California and Seattle late at night, and for much of that time there were no other cars near me. Just the occasional faster car that comes from over the horizon behind me, passes me, and disappears over the forward horizon, or the occasional situation where I'm the faster car doing that to someone else.
With an AI driver I could have read or slept most of the trip with the AI only asking me to pay attention when we are encountering other vehicles or there is anything unexpected near the road, and probably covered 99% of the cases where something happens that needs an attentive human driver to take over for.
At least they are being honest and not pretending it is about safety or consumer benefit.
> We will not sit by as bureaucrats side with tech companies, trading our safety [...] for increased corporate profits,
Did you skip over that part of the sentence?
The factory could also cut costs by employing child labour to crawl among the machines fixing breakages at risk of their limbs.
That's capitalism for you. Are you really surprised people opposed this?
Perhaps if you'd done the same you wouldn't have such a limited view of the world?
You were taken in by the selective use of quotation vs summarization.
Whether or not safety drivers are better or required is something that depends on the technology and changes as it evolves. This is the kind of thing regulatory agencies are made for.
More accurately - Jaron Lanier’s views are playing out, which got picked up by Yang somehow.
Right now, Waymo is probably the only autonomous operator with a good enough record to be trusted with heavy truck operations. Waymo announced back in July 2023 that they were abandoning truck testing for now, to focus on ride-hailing.[1]
[1] https://waymo.com/blog/2023/07/doubling-down-on-waymo-one.ht...
If we want to automate stuff, start with the easy stuff.
I fear that unless the penalty for negligence is severe we will see a lot of bad situations.
>I'm convinced that autonomous trucks are the way to go. Free from fatigue, bad training, bad driving, and all the other things that come with human drivers, it could save lives, and maybe even reduce cost for everyone.
This is peak HN techno-utopian thinking, that is only poassible due to a rather shallow analysis of the state of things: first, its myopic to rely 100% on FSD when they are currently colliding into things in cities that have favourable conditions for it (Bay Area, AZ etc...). Second, most of you probably do not drive much in middle America, the roads are often not marked and signage isn't entirely better.
Sure there may be a way to overcome all of these short-comings in time when ML training becomes more refined and can do much more capable than it can now. But then you have a severe lack of connectivity in much of the US (hence why Starlink is such a massive game changer) and before you say 'well then just make them use starlink) I think Elon has proven himself to be unreliable in ensuring he can deliver services without some reason that only make sense to his deluded mind being a real threat.
Honestly, if people want to take these kind of risks on the road then they need to at least be well informed on the matter: drive through the Rockies Mountains during an avalanche or even a blizzard, hell drive through Texas during one of their snow storms that takes out the entire grid--I've done all 3. You'll see why contingencies need to be in place, because modern logistics is a modern miracle and that relies on a myriad of moving pieces all working toward the same goal and even then things break.
I'd be open to the idea of those Human piloted convoy methods, with a Human pilot to correct where needed, until things get to that miraculous inflection point where FSD is possible but we are so far away from that this needs to be taken serious because too much capital has been destroyed on things that quite likely will never come to fruition, and FSD being around the corner is one of them.
I'm convinced that automated trains are the way to go.
Way more energy efficient per cargo pound mile. Technology already exists. Can easily ge electrified without expensive batteries that need replacing every few years.
"AI" Piloted trucks are not.
Truck freight is already much cheaper per lb/mile than truck freight.
After you roll in the costs to develop, build and maintain a 500k-1million dollar self driving truck it only gets worse.
I genuinely don't think I have the diction to adequately express the inanity of a policy requiring humans to babysit machines and the resulting loss of economic efficiency. Like I said, it is something I would expect to read in a dystopia, about how shitty and unfulfilling life is that the government needs to invent make-work jobs like sitting in automated trucks. Or in a history book about how some once-great empire has fallen "the consuls in their wisdom dictated that every slave must have an overseer, who is required to not work an only observe the work of the slave such that the working slave not accidentally deviate from his or her assigned tasks -- the obviously resulting crash in grain output was one of the contributing factors in the destabilization of the 8th dynasty".
I am also reminded of wikipedia article I read recently, from an HN post, about how the Soviet union starved themselves (and china followed) because they put a psuedo-scientist in charge of their agriculture.
>new automation
I missed the part where this would've been a transitory policy with a clearly defined and reasonable exit condition? If so I judged too hastily, but I doubt it.
Also newspeak is about removing words and simplifying the language. If anybody talk about 'newspeak' because new words were added, they're lying about having read the book or morons, I don't see a third option.
And you don't think the veto translates to something along the lines of 'corrupt regulatory capture for the benefit of tech companies'? Just because you don't like the proposed solution doesn't make it double speak for something you think.
> where the government uses double-speak to convince the people
It was used to control the narrative/population, not convince people - everyone knew it's obvious BS. But there the government was implementing totalitarian system squashing opposition. Here, we've got agreement from majority of both the government and population and a veto from a single person higher up. It's almost the complete opposite of the book.
> how shitty and unfulfilling life is that the government needs to invent make-work jobs like sitting in automated trucks
Wait until you hear about mining and farming subsidies in lots of countries and how much voting weight those groups had around the time of increasing automation and phasing out coal. Or why corn is in everything in the US...
How precisely is not forcing an industry into an inefficient[0] mode of operation 'corrupt regulatory capture'? It's pretty literally the opposite of regulatory capture.
(hence when removing the not the above sentence becomes regulatory capture, regardless of how double-speakily it is described).
[0]A non-market derived solution.
What I meant though is that the veto may be simple opposition to the idea, or it may come from having a special interests of tech corps influencing that decision - given popular acceptance of the proposed law, that would make it corrupt regulatory capture.