Hacker News Reader

Comment by DebtDeflation | Hacker News Reader

No, the entire promise was to eventually have a fleet of autonomous vehicles doing the driving. The human driver phase was just to 1) drive adoption of the app and grow share and 2) generate training data for the self-driving AI models while waiting for the fully autonomous capability to arrive. Scale is irrelevant as long as you are paying humans to drive cars, because that's not a model that can ever scale and will always just be a cab company regardless of size.

View on HN

>No, the entire promise was to eventually have a fleet of autonomous vehicles doing the driving.

No, to the best of my knowledge, the push towards autonomy came after they realized that they would struggle to get profitability through scale. Just look at their first pitch deck[1] from 2008, there's no mention of autonomy. They didn't really start pursuing autonomy until around 2015, when they finally started hiring researchers for that[2].

[1]https://techcrunch.com/gallery/here-is-ubers-first-pitch-dec...

[2]https://archive.is/pklu6

I’m not sure if I ever believed that hogwash. I think it was a fantasy meant to try to justify ludicrous stock prices on unprofitable companies. After 14 years of serious investment, it’s not clear if self driving cars are ever going to become mainstream, and certainly not soon enough to keep Uber and Lyft afloat.

It's a bit odd to be skeptical that self driving cars won't ever be mainstream. Computers have only been a thing for ~78 years and yet they're already capable of doing so much.

But now instead of drivers, you need to acquire land/buildings/people to manage the fleet of self driving cars. Someone has to refuel/recharge the vehicles, clean/maintain them, etc. Sounds like a taxi company to me.

That's a fair point. Maybe self-driving cars aren't practical at scale but it seems like they'll still be a thing.

Well, that was their initial idea, but I think they figured out a few years ago that it wasn't going to happen in time, and tried to find a different way. In theory, software cost scales less than linearly with customer count, so increasing in scale would reduce it as a % of revenue.

Although, maybe not if all your money goes to AWS anyway...

But they don't sell software, they are a middleman selling labor (drivers). They were never going to have software product style scaling.

Is Uber genuinely gathering driver data? They're only managing a fleet, and it seems unlikely that each vehicle has cameras installed. Wouldn't the autonomous driving tech actually come from collaborating companies instead of Uber directly?

Reply on news.ycombinator.com