Hacker News Reader

Comment by dilap | Hacker News Reader

Depends on if Tesla succeeds or not. I'm betting they will.

I can't suppress the strong feeling that they will, not by being technologically mature, but by easing a lot of regulations that are restraining them now.

Especially with Tesla's Ceo soon to be a member of the government.

Just like cars companies made jaywalking illegal or bought public transport companies to close them.

Yes. I expect that self-driving Tesla accidents will be rapidly suppressed.

Is Tesla a serious competitor in this space? I know they've made a lot of noise about self-driving cars, but that all seems to be hype generated to sell more human-driven cars. Maybe I'm off base here. I don't really follow Tesla closely.

Tesla's announced Cybercab literally can't be driven by humans, it has no steering wheel/yoke.

https://en.wikipedia.org/wiki/Tesla_Cybercab

From your link, the Cybercab is a concept with an unknown release date using technology wasn't even demo-able during the reveal. This seems more in line with my comment about generating hype rather than a convincing show of capability. Even if they do release it, which would be cool, wouldn't it be like a decade behind Waymo at that point?

No, Tesla has let select users run full self-driving betas for years, it's a very capable system now.

Here's a good and relatively recent (3 months old) video about how its current capacities:

https://www.youtube.com/watch?v=6qY51q2Zifc

If you want to see how the latest beta drives, it's version 12.5.6: https://www.youtube.com/results?search_query=tesla+12.5.6 .

It's still far from ready to release, but it is also improving very quickly - far from solely hype. They have the most impressive non-geolocked self-driving system as far as I'm aware.

It depends on what you mean by competitor. afaik Mercedez sells a self-driven level 3 autonomous car where they even insure the driver/car, but it's pretty expensive. Tesla has the numbers, but not a single car that can do as much as the Mercedez. Tesla promises they can just push an update an everybody gets a self-driving car, but they've been saying that for... 7 years maybe?

It still won't be a big bang moment. There might be a moment you can go hands off on highways, and a moment you can go hands off in urban environments. But it's still going to be gradual overall.

I've been a bit surprised that no one (well GM was mildly pushing in this direction at one point) has really gone after autonomous driving on selected highways within some weather parameters. I know urbanites get all worked up about robotaxis but a system that lets me chill for the next 100 miles of highway driving seems far more interesting if it's reliable/safe.

Highways is this weird thing where building the L2/L3 capability is comparatively easy (Which is why Tesla autopilot has been halfway decent for a long time now, and BMW have a true L3 system in restricted areas where you can take eyes off the road), but L4 is really hard because a true L4 system has to always be safe even if the driver doesn't take over. You could be tens of miles between highway exits, and just stopping on a highway isn't exactly safe by default. An L4 system therefore needs to handle anything between its current position and the next exit which massively ramps up the number of possible situations it needs to be able to handle compared to the L2/L3 equivalent.

L4 in an urban environment is weirdly feasible by comparison, even though we still don't have examples of extremely good L2/L3 systems yet (there's far more complexity by default compared to highways: lights, pedestrians, complex lane decisions, more visual noise, higher average relative speeds) we already have these L4 Taxis being rolled out. The reason is that if a driver doesn't take over it's much easier to find a "safe" place to stop. Apart from getting stuck in the middle of a junction, you can feasibly park up anywhere that has speeds <=40mph.

My understanding is that Tesla's approach is to not require hyper-detailed maps. So far it doesn't work well enough, but if they do manage to get it working, then I think it will actually be a "boom, suddenly it works everywhere" moment. Waymo One, by contrast, requires extremely detailed maps of the service area, which is why they can only slowly roll it out place by place.

It's not about maps, it's about operational design domain.

The requirements for a self-driving vehicle on highways is very different from urban and and very different from sub-urban. Multiply that with how different countries and cities have very different looking roads, e.g. narrow windy roads in London vs wide straight roads in Phoenix.

I agree it's not going to be an location-by-location rollout (other than for regulatory reasons). But I do think Tesla will find that their product "just works" in some places, and doesn't work at all in others - at least for a time. I don't think Tesla would sit on their hands and say "it's not ready yet" until they'd cracked everything. If it's safe to take a nap on Highways and not safe in an urban area then I think they'll just start selling that.

Reply on news.ycombinator.com