Uber Sells Its Self-Driving Unit To Aurora

Uber's self-driving unit, Advanced Technologies Group (ATG), is being acquired by its start-up competitor Aurora Innovation, the companies announced Monday. From a report: The deal, expected to close in the first quarter of 2021, values ATG at approximately $4 billion. The unit was valued at $7.25 billion in Apr. 2019 when Softbank, Denso and Toyota took a stake. Uber CEO Dara Khosrowshahi will join the company's board, and the ride-sharing giant will invest $400 million into the company. Overall, Uber and ATG investors and employees are expected to own a 40% stake in Aurora, according to a regulatory filing accompanying the deal; Uber alone will hold a 26% stake. The start-up is being valued at $10 billion in the transaction, according to a person familiar with the terms of the deal.

Is this the end for Uber?

[Way, Way Smarter!]

Uber has never made a profit. Uber can only make a profit if the drivers no longer need to be paid -- ie. self-driving cars.

It's possible that Uber could license the technology in the future. It does make it easier for Uber to select a self-driving solution from competing vendors, when such solutions actually exist in a cost-effective manner.

Otherwise, this looks like a short-term grab for cash at the expense of long-term profit.

Re:

[S_Stout]

Looks like a short term cash grab to keep the ship from sinking. I think Uber has done the math of how much money they have currently, how much more investors are willing to give them, how much they're currently bleeding cash, and how far away we are from true self-driving cars. Uber isn't going to survive long term. But, their top execs can still milk this cow for all it's worth before investors realize it.

Re:

[Mitreya]

Otherwise, this looks like a short-term grab for cash at the expense of long-term profit.

I am shocked... Shocked that anyone would do that. :)
The entire Uber operation seems like a short-term grab for cash that has gone on for a surprisingly long time (at the expense of investors and drivers).

Re:

[Gavagai80]

Uber's self-driving was a decade behind everybody else, already killed someone, and going nowhere. There's zero chance of their product being the first to market, so owning it would merely prevent them from being able to license and use self-driving in their fleet for many unnecessary years. Finding a sucker willing to pay so much for something that has essentially negative value to your company is a great decision.

Re:

[jeremyp]

I don't think that self driving cars are Uber's saviour. They'd have to make an enormous capital investment to buy and maintain self driving cars which is completely counter to their business model. They'll never be able to persuade other people who own self driving cars to rent them out for Uber journeys, at least not after the first viral picture of a car that came back to its owner after one of the passengers took a dump on the back seat.

Smart Move

[GregMmm]

Is the idea good, yes.

Would it be used, yes.

Who in the right mind will take on the financial costs of all the law suites when (not if) there is a problem. Maybe someone will make lots of money, but there will be "bumps in the road" they will have to deal with.

Re:

[subreality]

Who in the right mind will take on the financial costs of all the law suites when (not if) there is a problem.

Insurance companies. Evaluating risk and charging appropriate premiums to cover it and keep a slice for themselves is what they do. They've got this.

Translation:

[Rick Schumann]

Uber, seeing that so-called 'self driving cars' are a dead-end and will never be anywhere near 'autonomous', and having come to the game late, offloads the White Elephant that it is on some unsuspecting suckers who still actually believe that magic can happen.

Re:

[Pascoea]

Never is a long time. Given how fast tech has been growing, even in my relatively short lifetime, there is very little that I would put in the "never going to happen" category. There is nothing magical about operating a motor vehicle safely. Unless there's some sort of ESP that's helping you drive there's nothing that can't eventually be replicated with sensors, computational horsepower, and proper logic. There's just argument whether "eventually" is now, next year, 5 years from now, or 50. It's not "

Re:

[Rick Schumann]

*sigh* THIS again.
The so-called 'AI' has ZERO capacity to THINK. Therefore it cannot replace a human driver. Ever. We need General AI not 'deep learning algorithms'. It just plain is not sufficient and never will be. If it was so goddamned simple they would have them already everywhere but they cannot get it across the finish line because all the 'training data' in the world is not enough.

Re:

[Rick Schumann]

We do not understand how 'thinking' works therefore we cannot build machines that 'think', and 'thinking' is what is necessary to properly drive a vehicle on open roads. 'Deep learning alogorithms' won't cut it, never have, never will, too limited.

Re:

[Pascoea]

The road of progress is littered with people who said "it can't be done", passed by those who had the vision and drive (pardon the pun) to go get it done. You don't need a computer that can "think" the way you and I do. You don't need a machine that can paint pictures, design bridges, compose music, or any of the millions of wonderful things the human brain can do. A self driving car needs to know where it is, where it's going, and what's 100 feet in front of it with an incredible degree of accuracy. The se

Externalizing Liabilities

[eepok]

Uber's smart to remove such a massive risk from its balance sheets. They will be closer to profitable (at least in the short term) and still keep a 26% stake in Aurora and affiliate investors keeping 14%. That's enough to steer Aurora and possibly re-absorb it if/when they ever nail down an AV package. And if Aurora kills people with its AV (like Uber has), Uber won't be footing the bill. They could possibly lose their investment, but they will keep their shirts.

There is some funny math going on

[magzteel]

There is some funny math going on. I think this is funky transaction to puff up both companies balance sheets.

Haha, uber.

[groobly]

Haha Uber. Not as easy as it looks, is it? AI has been overhyped for... let's see... over 50 years now. Maybe exceeded only by nuclear fusion.

cut cut cut

[recharged95]

Uber also getting out the flying car biz [bloomberg.com]