Hacker News Reader

Comment by Preemo | Hacker News Reader

Is anyone else getting tired of Uber as a whole? There's nothing amazing about the company itself, they've hardly been innovative beyond the idea of ride-sharing. Can't wait for Uber to crumble from the top down starting with their over-zealous, dark-triad CEO.

View on HN

Not to mention their business model appears unsustainable to begin with:

Part 1: http://www.nakedcapitalism.com/2016/11/can-uber-ever-deliver...

Part 2: http://www.nakedcapitalism.com/2016/12/can-uber-ever-deliver...

Part 3: http://www.nakedcapitalism.com/2016/12/can-uber-ever-deliver...

Part 4: http://www.nakedcapitalism.com/2016/12/can-uber-ever-deliver...

Part 5: http://www.nakedcapitalism.com/2016/12/can-uber-ever-deliver...

Part 6: http://www.nakedcapitalism.com/2017/01/can-uber-ever-deliver...

Part 7: http://www.nakedcapitalism.com/2017/01/can-uber-ever-deliver...

Business model is fine it's just that the problem that arises from the wanted business model is hard. Given their bad reputation, especially now, they just don't have quality engineers to solve the problem.

Ride sharing is computationally NP-hard (assuming vehicle capacity larger than 2), the whole fleet management itself is NP-hard and depends highly on the realtime routing capabilities and prediction.

To make a combinatorial optimization algorithm capable of incorporating ride-sharing constraints and realtime data is a serious feat which maybe they jumped in too early to solve (given that no current research published tries anything similar). For really groundbreaking work to happen one needs quality engineers and researchers, and realistically, who would want to work in a company that is openly criticized for having terrible work culture?

The only way they can lower the prices and then find the sweetspot prices after monopoly is by having superior machine learning and combinatorial optimization solution.

I don't think this problem is as hard as you think it is in practical terms. N is going to be relatively small given basic constraints (the number of drivers and requests that could feasibly be matched is geographically limited). The search space for a request is limited to a few hundred drivers at most, competing with what is likely an even smaller pool of requests, making even extremely inefficient brute force approaches feasible.

Yeah, problem is so simple that they have job listings for AI research engineers that should have a little bit of experience in TensorFlow, Theano, Caffe or Torch (obviously some lovely deep learning on who knows what). About a year ago they were searching for PhDs and asked for experience with combinatorial algorithms, especially travelling salesman and vehicle routing problems.

It's worth pointing out that Uber had their service working more than a year ago, so while the folks you mention might help take Uber to the N+1th order of optimization they are clearly not necessary to reach the Nth.

I suspect this is what the GP is getting at: reaching the levels of optimization on this problem which are required to launch the service is comparatively easy. Going from there to optimality is extremely hard, but may not be necessary from a business or end user point of view.

N that the GP is talking about is number of vehicles and number of requests.

Yes, service can obviously work with subpar algorithms but to really succeed at pricing it as cheaply as possible it requires practically an ability to successfully predict the whole day and then optimizing on the NP-hard problem of that whole day. Maybe sampling a million day variants while routing to make a single decision (which driver should pickup the next request).

Of course brute force greedy algorithms work but they can be, on a hundred vehicle scale, 30% away from the optimum cost.

I've been downvoted to oblivion so I cannot longer keep participating in this discussion (HN will shadowban me).

I've been experiencing something similar in my own life.

It comes down to basically believing what I immediately see and experience, rather than or regardless of: 1) What I am told; 2) What I might hope for -- either immediately or "at some point in the future".

There is a saying -- an aphorism? -- for this, that applies particularly -- but not only -- to initial impressions. (I won't say "first", in case of a literal bad day for someone.)

When someone shows you who [or what] they are, believe them.

This is also phrased as:

When someone tells you who [or what] they are, believe them.

When I reflect back on experiences with people who have ultimately disappointed me, I often find that indeed, they were telling me more or less directly that this would happen.

In descriptions of themselves. In descriptions of their other relationships. Sometimes in a particularly poignant word or sentence. It was there -- they were telling me so, right at the start.

I find that for institutions, too. And the personalities that comprise them and guide them.

Hell, in this case, even their name choice: U[e]ber. Really?

I'm getting tired of Uber. And AirBnB too for that matter. I don't know anything about the company, but the business model and rediculous valuation seems pretty much the same.

Reply on news.ycombinator.com