Current language models (misnomered as “AI”) are great at some things but fail at any real thinking.
Self-Driving works decently in predictable environments but anything outside of those limits can make it literally crash and burn.
Public road transportation by individuals just has too many cases where real decision-making is required.
Despite all their resources, I think they’ve given up. All the brilliant engineers and scientists have given up, because they know what we’ve suspected for a long time.
Eh, Musk’s pivoting a bit towards humanoid robots. Self-driving cars are sorta off the table, and Musk’s car brand in general is not doing well due to his political shenanigans and the tone of US international relations recently.
It’s much more beneficial to him and the ultra elite, to develop humanoid robots that can get deployed as a civilian pacification paramilitary force, anywhere in the world, supported by Starlink etc. With enough investment/production, a very small ultra wealthy elite can completely control the world in a very overt, unapologetic and totally unaccountable way.
Like if a paramilitary robot screws up and shoots some extra protestors, it prolly wouldn’t even make the news, which’d be controlled by similar billionaires with their own robot paramilitary forces. But a self driving car screws up and causes accidents, and you get all these poor people whining at you, with no recourse to just have your robots shoot them silent.
If they’re depending on musk for this tho, they’re fucked. His stupid ass robots are vaporware.
Self driving cars have to be the most roundabout way to avoid railways that I can think of. It’s a stupid invention developed by greedy car brained idiots whose efforts are totally outmatched by any modern train system in any developed country.
I mean we should absolutely build more railways and we should absolutely have more bus routes, but there are times when a car makes the most sense.
Especially in more rural areas. If I want to visit my parents I can take a short walk to a bus or rail to a different bus or rail, but then I’m stuck. There isn’t and there will never be a bus or rail that goes anywhere near their home because it’s just too far in the middle of nowhere (and honestly it’s not really that far in the middle of nowhere).
Not to mention that the US was designed for cars. Arguably that’s where a lot of these problems started, but you can’t just be Captain Hindsight and demand we rebuild absolutely everything.
We should absolutely be investing and funding rails and buses much more than we are now. It’s embarrassing how far behind the US is.
Then there’s even less reason to invest in self driving right now if we’re aiming to minimize the amount cars
Edit: wording
This post caused me to look up what changed: sadly they aren’t exiting the consumer space but rather are just ceasing to include lane-keeping in their basic package to instead require new customers who want it to pay a monthly subscription fee for what they amusingly call “Full Self Driving (Supervised)”.
Yup. Self-driving cars will never truly be possible with unpredictable humans still driving the other cars on the road. Maybe if you could put a true AGI into a car.
We could have self-driving cars with today’s technology if the only other vehicles on the road were also self-driving and all interconnected via a mesh network with each other and the other critical road infrastructure like intersections.
But if on today’s roads if you want a truly Level 5 self-driving car where you get in the car in your driveway and say “take me to Burger King”, you’ll need an AGI in the car.
A train is so much easier if you don’t want to drive. But billionaires don’t want trains.
The only way self-driving vehicles can work is if roadways are designed for them.
Like how they can sometimes work as shuttle vehicles for airports parking lots and such if they have designated roads that only they use.
Human perception took literally millions of years to develop to interpret and interact with the world. And it requires the capacity to understand.
AI (probability prediction models) can’t do that. And they never will.
I still remember redditors telling me with supreme confidence in 2015 that there would be self driving cars in five years, all truckers would be unemployed, and there would be no more traffic accidents.
I bet they don’t remember that though.
Back then, I figured we’d have self-driving cars and boots on Mars by now. My perception of the world sure has changed a lot since then.
heck I thought we would have self driving cars by 2015 at one point.
They are looking down at you, laughing from Mars.
Hope they enjoy the oxide-filled soil and the lack of air.
They are exiting due to musk shooting his mouth off and not being to really compete with anyone.
He wants to build robots so he can end up taking cats with opposable thumbs to Mars.
Its a much more likely scenario than him getting there.
It has hit the limits because moron Musk insists on Teslas using only camera, and not camera + radar and laser, like other successful self-driving cars do.
You don’t even need cameras, you need rails on the ground you know, car follows the rails. Also a driver maybe, if you have a driver it should work, you tell the driver where you want to go, they take you there.
Otherwise, you also can try a global satellite mesh, connected to laser squirrels that can tell the satellites what to follow, then it tells you have how to drive. But you need at least a nuclear power plant to run your setup, 59 giga server farms that are on high pressure, high fule mode.
Maybe, but the other guys with lidar are using 1.8 RDP’d in Filipino drivers (without US driving licenses) per car… so maybe its a pipe dream either way.
not camera + radar and laser, like other successful self-driving cars do.
camera + radar and laser …and guys in the Phillipines.
Ditching lidar was madness, at least if you care about safety. Musk might’ve had a few brain cells left at one point, but the nonstop dose of horse‑tranquilizer is clearly wearing him out. He and his pals love llms because they see themselves in those bots: great at spitting out slick word chains, but with zero real understanding of anything.
Musk not going to Mars anymore. Musk exiting Tesla consumer space. Ha! Ha! Haaaa! The inevitable collapse of his entire house of cards has begun! Soon he will join the fent zombies he so despised, leaning in the street!
Self-Driving works decently in predictable environments but anything outside of those limits can make it literally crash and burn.
Public road transportation by individuals just has too many cases where real decision-making is required.
Despite all their resources, I think they’ve given up. All the brilliant engineers and scientists have given up, because they know what we’ve suspected for a long time.
I mean, Waymo is still actively expanding to new cities and expanding its coverage areas within cities.
Tesla may have given up on the consumer market for that reason if they were banking on a super llm based self driving system, or it could just be because they got burned when Elon rallied consumers against them.
Supposedly Musky is going to build robots instead of cars.
not supposed, he shut down two Teslur models and is tooling to make sexbots.
Onwards to the next grift!
he build taxis that runs people over.
Right after we colonize Mars. Any day now!
to be fair with ketamine, elons musks brain is already there.
Next year!
Hold on, wait, whoa, really? The great and respectable Tesla has to exit the consumer space? What? No! Surely big brain genuis can find a solution! They can’t? Wow. I am sure each and every one of you will have wet sloppy pillows tonight, from all the crying! I am shocked! Shocked I say!!!
…actually, not that shocked.
elon is probably hopefully to leech of govt contracts ever.
Elon won’t give up, he can’t fathom not being right about something.
Usually far right about something
You lost me at “misnomered as AI”. Artificial intelligence as a branch of Computer Science is a broad field covering not only LLMs - themselves a sub-branch of Deep Learning - but also Fuzzy Logic, Bayesian Reasoning and other statistical methods.
By any canonical definition; LLMs very much are AI.
Perhaps you meant ‘AGI’ as in Artificial General Intelligence which means ‘human level across every domain’. Or, perhaps you know this and we’re just being snarky about their current capabilities.
They’re saying that people are now referring to AI as LLMs only.
Misnomer: “an adjective used to describe a person, place, or thing that has been given an incorrect, inappropriate, or unsuitable name or designation”
It’s like when you keep calling all squares rectangles. Yes it is technically correct, however, its an unsuitable name. Its misleading.
I would also add to this and highlight that the average person does not differentiate between AI and AGI - I’d bet most haven’t even heard of AGI
Tesla sunset two of their models, three still for sale and being built. Definitely not giving up on AI or FSD, they are doubling down on cybercab, and elons pay package is tied to FSD subscriptions. If anything, the messaging is they are ramping up AI investment. Humans crash and burn regularly too. Brilliant engineers have definitely not given up, I’m not sure why anyone would think that. AI has continued to improve pretty steadily in all areas (see benchmarks)
