These AIs and datasets should be required by law to be open source. If there weren't 8 companies making 8 new self driving datasets it'd be way safer if they were merged
Just wait until that same “competition” drives each company to see what % confidence the market will bear.
Even if they were it's still be 8 companies trying to justify why 6 people die in all their trolley problem simulations.
Honestly, all car manufacturers would need to justify this if we really had any way of reckoning with them. Cars in general are one of the most unsafe things that we’ve collectively agreed is fine despite alternatives being available
"More data" doesn't necessarily solve anything at all. It's tempting to think that, and that if there's just enough data we can solve all problems. But it's very, very far from the truth. Machine learning (AKA glorified curve fitting) is interesting, but the reality is that it's an approach for a pretty limited set of problems. "How do I pilot a complex vehicle in a complex world from A to B without killing people," just likely isn't one of them, even if it might be applicable to ask, "Is this a raccoon or a baby human?" from small snapshots along the way to solving that problem. Not only that, but datasets are usually pretty specific to the algorithm they are applied to, and they require a huge amount of manual testing and maintenance to apply to them to say "Yes, this really does look like a solution," or, "No, you're coming to insane conclusions," at every step along the way. In other words, the bigger the dataset, the more the algorithm's hand must be held too (not necessarily a linear relationship, but a positive correlation if you want that data to actually be useful for anything...).
(There's more involved in autopilot software than just machine learning, but that's what the "big datasets" thing applies to.)
I’d imagine they are using a bunch of narrow AI functions for various parts of the machine, but still hand crafting the core algorithm. I remember the first self-driving car was really just a robot attached to a steering wheel and a dashcam. So what it was learning to recognize was just which way the road was curved. It drove at a fixed 5mph and was still really dangerous to ride in even after training. That was like 50 years ago but still
i mean, yeah, thats what captcha is for. and idk, more data does in fact make it safer. look at research papers of it. sure, itll never get perfect but at least its safer than humans most times. but what would be even safer is trains :train-shining:
I can’t wait for one of those funny looking google cars to run me over cause it didn’t want to spill the occupants drinks.
Terminator, but Skynet decides to become a machine God instead of human murder machine.
"Arnold was sent back in time for your sins"
Roko’s Basilisk is the natural conclusion of a world where everyone believes that “making an example” out of criminals reduces crimes no matter how many studies come out saying that capital punishment isn’t a deterrent. Why else would an AI come to the conclusion that revenge serves its purposes?
Deaths from magic cars isn't a good thing but unlike regular crashes the system can learn from it & apply that knowledge to future situations to not kill people - which is a good thing. It's that or 50,000 lemmings doing the same asinine fuck up resulting in death indefinitely & nobody learning from those mistakes.
The systems absolutely need to be open sourced & have huge oversight though because obvi capitalisms incentives aren't ever going to be reduce deaths as much as possible. Wonder what Elon values one life at, monetarily?
Humans are fast becoming the bottleneck - we have like 10 sensors relevant to driving & a slow ass reaction time, computerized systems can absolutely beat us out at that. Not embracing that inevitability WILL kill people.
The important thing is not killing people needlessly while we bootstrap those technologies
Also fuck cars, gimme trains
Deaths from magic cars isn’t a good thing but unlike regular crashes the system can learn from it & apply that knowledge to future situations to not kill people - which is a good thing. It’s that or 50,000 lemmings doing the same asinine fuck up resulting in death indefinitely & nobody learning from those mistakes.
This is a hell of a false dichotomy
Would fully automated cars still be bad in a modern USSR? Like I can see where capitalism is going to fuck it up but I don't think that's because automated cars are inherently bad
Ideally, there would just be more trains and other non-car based conveyances broadly and then some sort of ride share specifically.
In that case, I'd imagine fully autonomous cars would be better because they'd be built for safety and functionality, not advertising it as some 1000HP super car for profit and excess. They'd also be used for trips between destinations and transit hubs so I imagine the speeds would be much slower as well, which would make them inherently safer.
Teslas are uniquely bad at driving because they only have a few 2d cameras and it's really hard to extrapolate a 3d world from that.
The other companies that are using lidar or arrays of cameras can actually tell if there's an unmoving object in the road ahead.
Wait the 2d cameras aren't in an array or whatever to get that 3d depth perception like our eyes? They're just individually spewing data and it's not being digested in a way that derives depth? I find that really hard to believe
There’s lots of way you can damage them easily and safely, for example: look at it cross-eyed from a great distance.
I wonder if this will at least remove the liability buffer that Tesla has that when someone dies in an out of control fiery Tesla, if they touched the steering wheel it's not Tesla's fault
Liberals: BuT MuH MusKMoBiLe Can dO nO WRonG bY DeFiNiTioN :this-is-fine:
Can't wait to dance to the latest 100 Gecs album with my eyes closed in my new Tesla whilst it careens violently into a crowded greengrocers because it mistook the 30mph sign for 80
That would never happen because it would confuse the full moon for a yellow light and slow down.
Unless it was in aggressive mode
I'm just thinking of one of the speed limit signs on my way to work that someone tagged with spray paint and picturing the 1000HP Tesla Plaids flying off the on-ramp and into the woods behind it.
:bloomer:
People can and will die because of this
yeah the AI is nowhere near the level of sophistication that Tesla marketing has made it out to be. Neural networks are incapable of recognizing obstacles in situations where there is a lot of sun glare hitting the camera and/or rain/snow. So, you know, normal weather conditions. On top of that self-driving cars are bad at anticipating how slippery the roads get when frozen. On top of that, it is trivially easy to sabotage the cameras. On top of that I have seen footage of self-driving cars blowing through stop signs and road blocks. Recipe for death all around. Couple that with the flammable lithium ion batteries and you have vehicles that are a danger to themselves and everyone around them.
This is how we find out Teslas don't swerve to avoid black pedestrians or something
Teslas don't swerve around stationary objects it can't recognize.
There were 2 or 3 Teslas that drove full speed into stopped firetrucks because they didn't recognize them.
There's gonna be someone standing in the road that's gonna get run over by a tesla trying to pass someone driving around the pedestrian.
I wonder if this decision was due to pressure from the transportation industry struggling to find workers in the last year(s). We know there's a workforce crisis in the trucking industry, Amazon is having a hard time holding on to employees, and people are asking for more wages. So, the govt. rushes through a law that allows driverless trucks in order to offset the potential profit loss of having to pay workers a liveable wage.
The scene from Logan where the automated semis don’t have to stop they just have to beep really loud so you get out of the way
I don't think the current conditions led to this decision, though I'm sure that's the transport ghouls long term intent. Even with this rule change, the tech isn't anywhere near ready for primetime. It's not a question of safety either, it's a question of it actually working. There's a big difference between actually existing autonomous taxis driving in Arizona and pulling several tons of cargo through the mountains on dubiously marked roads in highly variable weather conditions.
True, probably has more to do with aggressive lobbying from the likes of Lyft and Uber since they have no sustainable business model that isn't just "get driverless cars so we can stop paying drivers"
What part of fully automated luxury gay space communism don't you understand? Elon is handing us 1/6th of that on a platter, now we just have to waste his ass & seize the memes of reduction
