Our road to self-driving vehicles | Hacker News Reader

Our road to self-driving vehicles

uber.com

38 pointssmaili18 comments

I love how they disingenuously imply that their end game is to have tons of self driving cars with a highly trained (and presumably highly paid) driver sitting in the driver's seat.

In reality, those drivers are there temporarily until the tech is good enough to phase them out (obviously). My guess is that the cars will be ready for full driverless before "driverless with backup driver" becomes big enough for anyone to notice.

I like asking my Uber drivers what they think about the future of self driving cars. I don't ask in a mean spirited way. Hardly any any drivers mention being phased out. I agree it's inevitable.

Those drivers will provide value add services. Which is more valuable: a coffee shop or a coffee shop with a person in it? Why would you think a mobile coffee shop would be any different?

Not sure this is the best analogy. Here, the car is the coffee.

So to use those nouns, it's more like comparing a cup of coffee vs a cup of coffee with a person in it....

Except they can’t afford them, and at some point even the most witless and puerile of VC’s has a chance of realizing that the level of automation required for profitability is decades away.

Curiously, the source code @1:44 [1] is pretty visible [2].

Seems a Flask app in a file called hexstub.py, created by dennisz.

[1] https://youtu.be/27OuOCeZmwI?t=104

[2] https://i.imgur.com/x4SJmJe.png

Neat. Looks like some source from their iOS app on the left, too.

We've had "Johnny cabs" for decades now in science fiction. People are clear on the concept. Once they work cleanly, smoothly, safely, I predict really driverless cabs will become the norm in half a decade. They need to look, feel, and be safe is all. Cabs don't have to handle all terrain; they can be constrained to known, plotted streets.

So, from the video, I get the idea that a "self-driving car" is going to still have someone behind the wheel (a more-qualified someone than otherwise, it seems) so the self-driving part to ... increase safety? make you good about automation? allow more videos to be made? ...

I was wondering that too. I think they're the designated scapegoat. If anything goes wrong, they get the blame.

Pretty smart. It sidesteps some of the moral questions by punting.

Or maybe it is to get people more familiar with the tech, and more "used to" it.

Exactly. We have this weird cultural fixation on having "someone" to blame. What's weird is that if it's a human that made the mistake, that's somehow OK. We give humans a free pass. "To err is human", after all. Shit happens.

But if an algorithm screws up and kills someone, that's unacceptable. We start imagining ways that a human would not have failed in the same scenario. We decry "killer robots" and "cold unfeeling machines". Even if that algorithm has already saved 99 lives from being lost to stupid, tired, drunk, inattentive humans, that 100th life is somehow more important.

I think that's already the play book. Remember when their self driving car ran 6 red lights in sf asnd they said:

>Uber spokesperson Chelsea Kohler said the car was being operated by its human driver at the time and had no passengers aboard

https://www.google.com/amp/s/www.wired.com/2016/12/ubers-sel...

Then it came this came out later

>Friday reported that traffic violations by the company’s self-driving cars were caused by problems with the cars’ mapping programs, and not, as the company had previously claimed, by human error.

After video of one of the violations surfaced in December, the company not only blamed and suspended the human monitoring the system, but doubled down. The supposedly human-caused red-light violation, they said, “is why we believe so much in making the roads safer by building self-driving Ubers.”

And while running red lights was the most egregious problem with the cars, they were also unable to safely navigate bike lanes.

https://www.google.com/amp/amp.timeinc.net/fortune/2017/02/2...

No, it's just that they left off the most disruptive part (for pr reasons). And that is that the people sitting in the seat are merely there temporarily until the software gets to a point that no human intervention is needed. And most likely, even before that happens, drivers will be replaced with a remote driver from a central office whenever a problem occurs that the ai can't solve.

They may be necessary initially but they also want to:

* Make people feel more comfortable with the idea

* Inspire the existing drivers; prevent a premature mutiny

It's a little sad that nobody at Uber spent 10 seconds reading past the first sentence in this post:

"At Uber, we’ve got lots of people helping to get our hundreds of our self-driving cars out on the road. "

I find this all very jarring.

I hear ya. I'm not worried. People tend to underestimate how intelligent humans are and drastically overestimate how "smart" pre-programmed cars are going to be. The reality is, like "free money" it's also a control system, and most of us already know it. Consider the near total rejection of camera tickets when it's put to a vote. There's billions of $ in hype. I'll happily bet against it.

Something that concerns me slightly more is when we start growing 'subservient' neurons to drive us around. They will be smart enough to make it work, but the closed and connected 'trusted' silicon surrounding them would be a problem.

A short term problem is new drivers might become dependent and not learn the necessary skills... like pilots forgetting how to land.... but I doubt it. Gen Z is turning out just fine.

Note I live in one of the most freedom appreciating places on Earth, if you are in one of the nanny states that has already hooked the population on public transportation then all bets are off and ya, you should worry.

Reply on news.ycombinator.com