Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tesla Robotaxi

tesla.com

Post reply on HN
Print view View original

481–490 of 788 posts

Re: Tesla Robotaxi

#481

Post by randerson »

throwaway48476 wrote: post #281
mavhc wrote: post #267

Earlier quoted context omitted.

Because getting GWs of power into one building and the cooling it is expensive. Why not let other people buy the computer, provide the power, and it won't need expensive cooling. Question is: are there enough high latency distributed workloads to sell?

High latency distributed with low security requirements. Until homomophic encryption works.

And resilient enough to handle nodes having an accident, catching fire in a flood, parking underground with no signal for days, pausing to be used as a car...

This is an unreliable spot instance at best, with none of the features one can normally attach to an instance (like storage, managed databases, ...). How fast can its Internet be? Will owners need to pay for Starlink too? (What about when parked indoors?)

It would have to be cheaper than all regular hosting options by a long shot for anyone to consider this. A very niche, low-paying market, in other words.

Re: Tesla Robotaxi

#482

Post by disillusioned »

f6v wrote: post #459
pavlov wrote: post #438

Earlier quoted context omitted.

Maybe in the US. It still does nothing in Europe, which is disappointing because the company happily took my money in February 2019 on the promise that this feature is imminent. I paid 7,500 euros for something that still doesn’t exist in any usable form. (I thought it would be good for my elderly parents who were the primary users of this car for a time.) It’s actively dangerous because it doesn’t even see speed lim…

> I paid 7,500 euros for something that still doesn’t exist in any usable form. And I thought people preordering games for 70€ were taking unnecessary risks. You guys took it to the whole new level.

For a period, FSD was up to $15,000 USD... For heretofore still unshipped vaporware. And I say this as someone paying the $100/month for the current very very beta, sometimes absolutely terrifying form of it.

Re: Tesla Robotaxi

#484

Post by jillesvangurp »

oblio wrote: post #402
jillesvangurp wrote: post #399

Earlier quoted context omitted.

Are normal taxis any different? Most of them only have one or two passengers. Typically in the backseat which gets kind of crowded with more people. You can get bigger taxis of course. But I'm talking about the common sedan model with the driver in the front. An unused front seat on the right. And the one or two passengers in the back. Exactly the same thing. Except there's no need for the driver. I don't get the mor…

> Are normal taxis any different? Most of them only have one or two passengers. Typically in the backseat which gets kind of crowded with more people. You can get bigger taxis of course. But I'm talking about the common sedan model with the driver in the front. An unused front seat on the right. And the one or two passengers in the back. Exactly the same thing. Except there's no need for the driver. We need to invest…

I don't think what you outline here, which I agree would be a good thing, and what Tesla announced are mutually exclusive. In fact, I think Tesla's announcement is just about the most concrete thing from any car manufacturer that gets us closer to that.

Getting rid of drivers means less space for parking is needed. It also means better utilization of the vehicles. So, less cars on the road. And they'll be electric so a lot more quiet and less smelly. And they also have a bus form factor; perfect for public transport.

Elon Musk actually showed some visuals of cities being greened this way. Complete with a cheesy joke about taking the 'ing' out of parking.

So, you are outraged because Elon Musk is doing exactly as you demand but just not in a way that you like?

Re: Tesla Robotaxi

#485

Post by benterix »

acchow wrote: post #421
wg0 wrote: post #416

Earlier quoted context omitted.

Elon has almost no credibility left for what he says. It's basically just a website at the moment with bunch of 3D renders which you too could get done from a web shop. Tesla has nothing new to offer and competition is catching up, EV adoption slowing down and such. If I had, I would gradually drop Tesla stock because it's going to go downhill if not rock bottom from here.

> It's basically just a website at the moment with bunch of 3D renders which you too could get done from a web shop. They had 2 dozen vehicles with no steering wheels taking attendees around the venue. The bar was staffed by Optimus robots. “Basically a website”?

> They had 2 dozen vehicles with no steering wheels taking attendees around the venue.

To quote a participant[0]:

> After over 10 years of Full Self-Driving development, @Tesla is limited to a 20-30 acre geofenced 5mph ride on a preprogrammed, premapped and heavily rehearsed route with no traffic and no pedestrians.

[0] https://x.com/realdanodowd/status/1844605093368512799

Re: Tesla Robotaxi

#486

Post by disillusioned »

jakeogh wrote: post #403
davedx wrote: post #314

Earlier quoted context omitted.

Also: > “The autonomous vehicles, we’re going to stop from operating on American roads, remember this,” Trump said.

Good. They don't work, and they wont work. We are a decade into the hype cycle on 'cars that have self' and the proponents are stuck pretending that their best case demo, which is always going to crash into stupid stuff that a human would easily understand (in it's highly tested and constrained route) has 'self' to preserve.

This is a reductive and far-too-pessimistic take when Waymo is completing more than 14,000 driverless trips per DAY and has completed over 15 million miles of driverless trips.

That's not a "best case demo": it's a real system, working in real world environments, and even if they're "constrained" (and go ahead and look at the Phoenix service area and tell me how constrained it really is), they are functioning, now, today. There's one driving by my house as I type this. It's not mowing down children. It's not causing fatal accidents. Its incident rate is dramatically lower than human drivers, and its severe injury rate is lower still. We're a decade into the hype cycle because of how hard a problem this is to solve, and we're finally catching up with the right confluence of technologies at the right time (mostly around machine learning, machine vision, sensor fusion, LIDAR, reinforcement training, and computer power) to make it actually work.

I can understand if you live outside a Waymo market that you might still believe this is still fairy tale "won't work" stuff, but when you live in a market where you see dozens of them every single day, doing their thing _unremarkably_, it's... well, it feels quite a bit like the future.

Note that I'm speaking explicitly of Waymo here. Tesla FSD still terrifies me, vision-only seems like a horrible oversight, pun intended, and while it's meant to be non-constrained, it still has a very, very far way to go to close the gap with Waymo.

Re: Tesla Robotaxi

#487

Post by A_D_E_P_T »

porbelm wrote: post #475
A_D_E_P_T wrote: post #465

Earlier quoted context omitted.

> Also it won't even stop .22LR bullets. Seriously? Is there proof of this? The body panels are pretty thick, and they're made of an austenitic nitrogen steel -- a modified form of 316LN -- which should perform nearly as well as any "armor grade" steel alloy. (Against steel fragments, it likely performs poorly, but against lead handgun rounds it is likely even superior to the average high-hardness armor steel.) Just…

There are videos of people shooting holes in it with 9mm at least.

I've seen a few where it stopped 9mm.

e.g., https://www.youtube.com/watch?v=vQ5EyKMqGGI

https://www.youtube.com/watch?v=hOteg0J5teg

Do you have links to those videos where it fails?

Besides, 9mm and .22lr are very different things. .22lr is one of the very weakest ballistic threats. 9mm has much more mass, usually more velocity, and a much higher (~3x higher) average KE load.

Re: Tesla Robotaxi

#488

Post by pavlov »

dom96 wrote: post #478
lowkey wrote: post #458

Earlier quoted context omitted.

- I believe he can launch rockets into space and land them on their own footprint. - I believe he can revolutionize auto manufacturing and disrupt a 100 year old industry replacing fossil-fuel burning dinosaurs with clean electric vehicles that outperform them and that appeal to the general public - I believe he can allow quadrapelegics to interact with the world in ways never thought possible - I believe he can, to…

You’re a fool tricked by another fool who shouts loudly that they support free speech while they ban speech left and right. For goodness sake, ElonJet was banned and you can’t even say the word “cisgender” on the platform. How delusional are you?

Also Musk has banned mentions of his own transgender daughter, who now posts on Meta’s Threads app instead.

X is like a textbook case of why total autocracy isn’t actually good management practice. Musk has become the Henry VIII of social media.

Re: Tesla Robotaxi

#489

Post by throwaway71271 »

Man I really have to stop reading hackernews comments, its a total bummer.

His companies are making rockets, autonomous humanoid robots and autonomous cars, I will cut my left pinky finger to work on any of those, and this is my capslock finger, so I will have to switch away from emacs..

With the worldlabs[1] work maybe they unlock models that do have even better spatial world model and can be used label data even better and faster, or create even better synthetic data so it can unlock FSD even sooner.

The fact that LLMs work means there is structure in language that is beyond our understanding, and yet the transformer can discover it and program itself to solve for it. I think that the stupendous amount of compute that is going to be released in 2025 will make it possible to train labelers that can do temporal labeling much better than humans and than the current models, and synthesize and perturb data to train really really good transformers that will outperform 90th percentile humans.

Why do you guys think that it can not happen? Maybe its not Tesla that does it, but I certainly think amazing tech is coming.

Hopefully it will not be just autonomous humanoid robots with guns paid by the military and patrolling the borders :astonished face:

[1]: https://www.youtube.com/watch?v=vIXfYFB7aBI (Fei-Fei Li and Justin Johnson)

Re: Tesla Robotaxi

#490

Post by Ukv »

fauigerzigerk wrote: post #449
pr337h4m wrote: post #330

Earlier quoted context omitted.

Humans don’t use lidar, which clearly shows that a vision-only robotaxi is very much feasible.

Humans don't act based on visual patterns alone though. We act based on our understanding of the world as a whole, including the intentions of other humans. For instance, when we see a ball rolling onto the street, we know that there is probably a young person nearby who wants that ball back. We don't have to be trained on the visual patterns of what might happen next. Of course AI can be trained on the visuals of hi…

> the number of things that can potentially happen is far greater than the number of training examples we could ever produce

Models don't need to have been trained on every single possibility - it's possible for them to generalize and interpolate/extrapolate.

But, even knowing that it's theoretically possible to drive at human-level with only the senses humans have, it does seem like it makes it unnecessarily difficult to limit the vehicle to just that. Forces solving hard tasks at/near 100% human-level, opposed to reaching 70% then making up for the shortcoming with extra information that humans don't have.

Post reply on HN