“Uber’s fare for a female from work to home after 10pm is higher” | Hacker News Reader

“Uber’s fare for a female from work to home after 10pm is higher”

twitter.com

23 pointsdevnonymous24 comments

I think that more than three data points are needed before screaming discrimination.

Exactly. It’s so frustrating that with some topics there is almost no difference between accusation and conviction.

I think it's interesting for discussion even if we don't seek to immediately condemn Uber for some wrongdoing.

If a computer breaks the law, and the company did not intend for that to happen, should the company be held liable? Should they be required to correct the program from infringing on civil rights, in this case, discriminating based on a protected class?

Is there any other evidence of this? Based on this tweet + replies, there isn't sufficient evidence to say gender is the most likely cause of the price disparity. For example, it seems at least equally likely that the price disparity is due to individualized pricing, based on personal usage history, rather than some blanket gender discrimination.

> For example, it seems at least equally likely that the price disparity is due to individualized pricing, based on personal usage history, rather than some blanket gender discrimination.

I think the point is that individualized pricing could lead to discriminatory pricing without any intent by developers or the company.

For example, maybe women take that trip more often than men. Then, on an individual basis, a woman might be more likely to be charged more, because she uses that trip more often.

This would suggest that in the future, complex machine learning systems might need to have civil rights protections built in, if we want to maintain these protections.

Uber employee here. No, absolutely not. We would not do this, and large swaths of the company, like myself, would quit if this were the case.

Could be a "innocent" mistake. Analysis shows that customers in cluster A are willing to pay 20% more than customers in cluster B for a certain route, as determined by machine learning. Turns out "A" means probably female customer, at night...

I wonder if there are any safeguards in place against this. E.g. take samples of male/female, different ethnicities, etc. that you want to treat identically, and check if they pay the same on average... or put that somehow in the pricing algorithm as a constraint.

I'm sure a naive algorithm with a lot of inputs would, after optimization and without supervision, make single women in bad parts of the city at night pay more. It will probably find a lot more cases to take advantage of, like white (black) people in a neighbourhood where they are the minority and don't feel safe or comfortable. Or rides from places with high percieved crime rates (it wouldn't know the crime rates, it would just know people pay more for certain routes).

This seems like the likely answer. Of course an unchecked algorithm would come to this conclusion, the key being that the algorithm is acting "without supervision". But it seems like negligence to not be supervising these algorithms.

How confident are you that you know everything that everyone at Uber is working on?

It is understandable for this to happen if there's any kind of machine learning deciding prices, it doesn't have to be designed discrimination to exist in this form.

Didn't a recent article describe Uber's extensive internal firewalls to prevent employees knowing what's going on?

Also, if you or other Uber employees would quit over this behavior, why haven't you already quit? This is arguably less egregious than many of the reported abuses of employees.

And yet, the evidence is there.

Will this be reported and investigated within Uber?

If every Twitter accusation and perceived injustice against Uber were investigated, there wouldn't be enough time in the day for anything else.

I had an Uber ride from a driver who accused Uber of paying a lower surge rate than what the passenger was paying. I assured him this wasn't the case but he insisted it. I told him exactly what I was paying, gave him my personal cell number and told him to call me if he collected less than this amount (less the Uber %). He never called me.

A lot of people have suspicions and conspiracy theories but Uber would not be in business if it did things like this. Most of the Uber employees believe strongly in what the company is doing, around the world, and exploiting women by charging them higher fees is not something we would tolerate. That's not how a sustainable company makes money.

> If every Twitter accusation and perceived injustice against Uber were investigated, there wouldn't be enough time in the day for anything else.

Perhaps having an opaque, complex pricing scheme isn't great for customer service or public perception of the company.

Personally, I agree and we will have to wait and see what happens ultimately.

Amazon tried this a few years ago, and I believe they stopped doing this but I'm not sure. But if I knew or suspected that Amazon was trying to squeeze a few percent from me, it would make me less inclined to use them. I can't see how this isn't the case with Uber as well, if customers suspect they are being gouged. It would then open the doors for vast accusations like the twitter post.

I'd also point out that swaths of people probably already have left Uber and tons of other companies for unethical practices.

You just don't notice because they get replaced, and the new people assume that existing practices are the norm. The world keeps spinning.

I guess my point is, if you don't see many people leaving at once, that isn't evidence there isn't something fishy going on. People have families, mortgages, children's tuition payments...

Perhaps, perhaps not. Uber seems determined to find out.

where is the evidence? OP didn't bother even posting a screenshot... This is just part of the FUD war going between old and new (taxi companies vs uber; old media portraying youtubers as far right extremists over a silly one-liner joke, retarded left-wing politicians taking selfies sitting on the floor of a private company train with dozens of empty seats around). best solution is not to get caught in the middle, but rather to sit down with a pack of peanuts and enjoy the show. the whole thing is actually really really funny as long as you now what it all is.

I find it hilarious you're trying to take the moral high ground in THIS particular case, which could be just chalked up as good business sense, and not in any of the immoral things they have actually done like stealing intellectual property via a shell company or sexual harassment in the workplace. All it really shows is your lack of economic understanding.

Businesses price discriminate. Against different bands of customers, perhaps by race, income, zip code or gender. It happens. It's smart. It's profit-maximizing. It's amoral. There's nothing inherently bad about it.

So what if a band of customers is being charged higher? What if they are actually willing to pay more? Why leave the surplus on the table? Are they satisfied with the service?

Look, I'm about as far as you can get from a free market Randroid, but this is just basic microeconomics.

This should have been a link to the Bloomberg article rather than an anecdotal tweet.

I wonder if it's possible to hack Uber's individual demand model by occasionally opening the app, picking a typical destination, and then closing as if it's too expensive.

Bloomberg article says Uber would like to price based on how much you are willing to pay instead of using some static factors like length and time of day.

https://www.bloomberg.com/news/articles/2017-05-19/uber-s-fu...

This is what every smart business likes to do. Charge more from those who are willing to pay, but also keep more price sensitive customers. Traditional way of doing this is to create multiple versions of the essentially same product and price them differently.

Are sure this is based on gender? Or just something that correlates to gender really closely?

Maybe the woman these tweets are from just use Uber more and so there also predicts they'd be willing to pay more. Or something like that.

Otoh, I wouldn't put it past Uber, but we only have a couple of data points here but even if we had a ton more, *my above hypothesis and actually discriminating by gender might be very hard to tell apart.

I'm pretty sure the code doesn't say

    if gender == 'f':
        price *= 1.3

But it is concievable that an algorithm found a cluster of people who are willing to pay more at a certain time, for certain routes. After all, that is what their rate finding algorithm does. That's what I'd make it do to maximize profit.

It just happens to be that the result is that some women (on average, or in some select cases, I don't know) are asked to pay a lot more at night. If it is based explicitly on gender, that would be really dumb, and I don't think even Uber is that evil. But if it is based on a close correlate, the effect is entirely similar.

We as a society need to think about if we want that, or if we want to add the constraint to such algorithms that men and women pay the same.

Is it OK that...

... people pay more at rush hour than at other times?

... people pay more in unpleasant weather?

... women pay more than men?

... people of different race pay differently?

... people of different income pay differently?

... potentially sick people pay more on the way to the doctor?

These are all things we have to consider. What do we leave to the market and algorithms to decide, and what do we constrain? Some cases are IMO absolutely clear cut, some we'll have to debate.

Does the company's intent matter if a computer program breaks some civil rights laws?

I see lawyers with dreams of dollar signs over such as-yet-undetermined disputes.

I feel as if I see claims like this pretty regularly. But when they don't amount to anything, I never see followups about delusions of persecution.

Reply on news.ycombinator.com