I imagine it is hard to determine how deep the water is. There is a lot of training with small puddles (ignore) and not much with deep water.
Still, it should be cautious as any human driver would be.
Waymo pauses Atlanta service as its robotaxis keep driving into floods
211–220 of 498 posts
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#212This is really my bear case against AI. I am not against it. I actually think it is really neat! But we have been working on driverless cars for how long and spent how much? And still things like a flooded roadway completely throw them. Tesla failed to deliver driverless cars but now is pivoting to the much more complex fully autonomous robots. And we can’t get AI to stop hallucinating facts, but any day we are going…
I'm not aware of any self-driving widely available ten years ago. I just took my Model Y over Highway 1 in California without requiring human intervention (other than when I chose to pull the car over).
Obviously when these things can become fully autonomous isn't absolutely clear, and there may always be some discomfort with a probability of failure without a human chain of responsibility.
But, given ten years ago this didn't exist at all for consumers, and it now more reliably does? It doesn't seem insane to think ten years from now, it might address more edge cases, and be safer and more effective.
Why would you look at the general trend and assert jettisoning the effort?
EDIT: It seems some of the tech started rolling out 2016; my mind mentally was thinking 2015. So maybe this started about a decade ago. Though still, the trajectory is a decade of these systems going from limited assists toward greater autonomy with demonstrable progress.
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#213This is really my bear case against AI. I am not against it. I actually think it is really neat! But we have been working on driverless cars for how long and spent how much? And still things like a flooded roadway completely throw them. Tesla failed to deliver driverless cars but now is pivoting to the much more complex fully autonomous robots. And we can’t get AI to stop hallucinating facts, but any day we are going…
AI as commonly discussed is just pretty-general intelligence that is very economically valuable. Not AGI outside of the true believers. And can we discuss AI drivers and AI LLMs in the same paragraph? One is a special application of trying to emulate a very particular human embodiment, with all the sensory challenges. The other is a brain in a vat. Both can fail and flourish independent of each other, or at least I s…
VLAs (vision-language-action models), which are offspring of LLMs, and their versions that are more suitable to edge devices are being used in self-driving to add common sense to path planners ("don't drive through a police standoff", things like that).
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#214This is really my bear case against AI. I am not against it. I actually think it is really neat! But we have been working on driverless cars for how long and spent how much? And still things like a flooded roadway completely throw them. Tesla failed to deliver driverless cars but now is pivoting to the much more complex fully autonomous robots. And we can’t get AI to stop hallucinating facts, but any day we are going…
The "responsible adults" know that chasing perfection gets you nowhere fast. A part of growing up is learning to put up with "good enough". A car that only fails in a road conditions edge case is good enough for the vast majority of cases. You accept that, and issue a manual override for when that edge case pops up. Then you add that edge case to your training sets. Then the issue never comes up again. If you think t…
Well its not that simple. In the same way that throwing an LLM into a process will always have a risk of blowing up spectacularly.
In this case it failed open. It didnt recognize that it was in an edge case (which itself is an edge case). So what are you proposing to be the solution to that? If the car itself does not recognize that its in an abnormal situation that needs intervention then how do you intervene?
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#215Earlier quoted context omitted.
I am a little worried that this is still a problem after 20 years. Don't they have simulators to test every weird and unexpected road condition offline? And flooded roads aren't exactly an unusual event to begin with.
In ATL this happens often enough that it's not a shock when it happens, we have lots of drainage problems here. I agree that I would have assumed Waymo had tested in events like this, but clearly not. So what I can say is running in ATL is a great test case for these events, and also the people who live here don't do a better job than Waymo did. There were dozens of people who ruined their cars yesterday trying to dr…
As much as one could expect waymo to train on it, one could also expect a functioning city to not have flooded streets
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#216This is really my bear case against AI. I am not against it. I actually think it is really neat! But we have been working on driverless cars for how long and spent how much? And still things like a flooded roadway completely throw them. Tesla failed to deliver driverless cars but now is pivoting to the much more complex fully autonomous robots. And we can’t get AI to stop hallucinating facts, but any day we are going…
The "responsible adults" know that chasing perfection gets you nowhere fast. A part of growing up is learning to put up with "good enough". A car that only fails in a road conditions edge case is good enough for the vast majority of cases. You accept that, and issue a manual override for when that edge case pops up. Then you add that edge case to your training sets. Then the issue never comes up again. If you think t…
> A part of growing up is learning to put up with "good enough".
"I feel content with good enough in this case." - quote from child whose body got folded in half by a Tesla
Your growing up and adulthood sounds a lot like settling for mediocrity from those who push shit on us without asking if we ever wanted it. Floods aren't a special edge case, they happen all the time. The people making these are so stupid and blind to reality they didn't think about the most basic 101 case of "what if it isn't a perfectly dry and sunny California day" because thinking isn't on the to-do list for these people. This shit is ass. Get it off the streets.
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#217Earlier quoted context omitted.
While this is going to be an overly optimistic scenario: Imagine how smooth a hurricane evacuation would go if _everyone_ used a self-driving car to do the evacuation - atleast there might be less gridlock than there is during any usual hurricane evacuations. And assuming the self driving cars don't do something stupid that causes every car behind it to essentially lock up and stop moving That said, I know a scenario…
Now imagine if the power is out and cell service is down. We saw that happen in San Francisco and it was chaos.
That's why on-board sensor only systems are the way to go.
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#218This is really my bear case against AI. I am not against it. I actually think it is really neat! But we have been working on driverless cars for how long and spent how much? And still things like a flooded roadway completely throw them. Tesla failed to deliver driverless cars but now is pivoting to the much more complex fully autonomous robots. And we can’t get AI to stop hallucinating facts, but any day we are going…
I will posit something that guides my own thinking about this; robotaxis will never drink and drive. I'll take whatever flavor of mistake they conjure over that. I can deal with stupidity, I cannot (and don't want to) deal with malice.
Many people don't drink and drive either. You can drive defensively, choose your own route.
Even on two lane roads: if an idiot overtakes into oncoming traffic there is usually just enough space for three vehicles next to each other. Can a Waymo move sharply to the right so there are two cars on each side with the overtaking idiot in the middle and all just fit on the road? I had to do that maneuver at least twice.
Can a Waymo prevent a carjacking when someone places traffic cones in front of it?
Can you open the Windows and get out if the thing decides to drive into a lake?
I don't know, currently defensive driving is the better option.
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#219This is a classic case of: if the situation is not in the training data then the model is unequipped to handle it.
We've seen the phenomenon before. We've been warned against the phenomenon before, and we'll see it again in other contexts in the future for sure.
Re: Waymo pauses Atlanta service as its robotaxis keep driving into floods
#220Earlier quoted context omitted.
The "responsible adults" know that chasing perfection gets you nowhere fast. A part of growing up is learning to put up with "good enough". A car that only fails in a road conditions edge case is good enough for the vast majority of cases. You accept that, and issue a manual override for when that edge case pops up. Then you add that edge case to your training sets. Then the issue never comes up again. If you think t…
> A part of growing up is learning to put up with "good enough". "I feel content with good enough in this case." - quote from child whose body got folded in half by a Tesla Your growing up and adulthood sounds a lot like settling for mediocrity from those who push shit on us without asking if we ever wanted it. Floods aren't a special edge case, they happen all the time. The people making these are so stupid and blin…
... which is different from the child whose body got folded in half by someone looking at their phone?
I think "good enough" ends up being okay. I _like_ driving. I would do manual mode often still just because I enjoy it. But I'd be completely fine with the option of autopilot in good conditions. Reality is that 99% of the time, my commute is boring and in good conditions. I don't need a self driving mode that can handle a blizzard when I'm in stop and go traffic and it's 20c outside.
This is much harder for Waymo since there isn't as easy of a manual override mode... But in my car? rip it.
Luckily I basically already have it. Adaptive cruise covers most of my cases well enough, but I wouldn't mind something with a bit more control (turning, etc.)