Live data from Hacker News

Google contractors reportedly targeted homeless people for facial recognition

theverge.com

101–110 of 158 posts

Re: Google contractors reportedly targeted homeless people for facial recognition

#101
post #68

I remember the outrage when people discovered that Google's AI wasn't properly trained on black faces. It makes sense that they try hard to avoid that happening again by paying black people to let Google scan their faces. It is not unethical to try to diversify your training data. https://www.telegraph.co.uk/technology/google/11710136/Googl... Anyway, this part sounds directly illegal, seems like it was just Randstad…

I feel like they were needlessly dishonest and misleading and that being truthful would have gotten more people on board. Saying something along the lines of "hey, we noticed our facial recognition algorithms don't work so well for African Americans. Could you help us fix that by letting us take a picture of your face? We'll give you a $5 gift card for your troubles."

That seems pretty clear. Verge title is kind of clickbait by using `Google Contractors` rather than `Randstad` though.

The "and/or" in "it sounds like Google and/or its contractor may have been taking some extreme and unsavory shortcuts to cash in." is clearly use of weasel words, the journalist was unable to substantiate the allegation that Google was aware of this unethical practice. Sloppy reporting demonstrating a lack of journalistic integrity.

Re: Google contractors reportedly targeted homeless people for facial recognition

#102
This seems like a fake outrage to me. If I were a homeless I'd be more than happy for someone to take a photo of me for $5. In fact, I'd find it pretty hypocritical if someone was spending their energy fighting against possible infringement of my rights in such scenario instead of actually providing me with money or food.

Re: Google contractors reportedly targeted homeless people for facial recognition

#103
post #69
post #38

Earlier quoted context omitted.

If you or anyone: - Provides a link or method to create the scan that takes just a few minutes (on Ubuntu) - Sends $5 to kauffj@gmail.com via PayPal or via BTC to 17h2GtaBzivnNtP24qoGg4a3pjgShkw7MD I will complete the process and post the result in this thread.

What's your point? You choose to do it because you don't care about giving your face picture to someone, fine. You're in a totally different position, financially independent I assume, you can weigh your options. A homeless person might not be in a situation like that. They're much more vulnerable and possibly willing to do things out of sheer desperation. Things they wouldn't do if they were in a different situation…

I don't know, you can do this to all sorts of scenarios. How many men are vulnerable to attractive women and will do all sorts of ridiculous things out of sheer desperation if a cute woman comes on to them? These men wouldn't do it in other situations like if they were already coming from a place of abundance, or if they had low T.

How many people in need of money do all sorts of desperate things like take a job they hate? How many HNers are working jobs they actually hate? Or they're being taken advantage of because they don't have the backbone to stand up for themselves? etc. etc.

Re: Google contractors reportedly targeted homeless people for facial recognition

#105

Earlier quoted context omitted.

But now they have a dataset where a disproportionate proportion (possibly the last majority) of people of color represented were homeless. That's bound to introduce other kinds of bias into the data.

Such as what? Some bizarre and unfounded idea that poverty and skin color have some correlation?

Such a correlation may or may not exist, but surely it's evident that no conclusion either way could be drawn from a dataset constructed using Google's method.

And yet imbalanced datasets are used all over the place, e.g. to identify "criminals" in China (https://www.newscientist.com/article/2114900-concerns-as-fac...) and the US (https://www.engadget.com/2019/08/14/aclu-facial-recognition-...)

Re: Google contractors reportedly targeted homeless people for facial recognition

#107
post #96
post #95

> Google and Randstad didn’t immediately reply to requests for comment. The content of the article is interesting enough, but this line at the end caught my attention. Is it reasonable to expect someone to "immediately reply" before you publish the article? Because that doesn't sound like ethical journalism to me, unless I'm misunderstanding the meaning of "immediately" in this context.

Doesn't matter, I suspect - It works within the narrative and implies they have something to hide. It's good /tabloid/ journalism, and poor investigative journalism. Randstad are very much the former.

Historically if you wait multiple days / weeks to give them a chance to reply they just do an end-run and publish puff pieces in major outlets to try and defuse your article. There are multiple cases of this in the past year.

It's either investigative journalism or it's not. How long you wait for comment has nothing to do with that. Do you really think this is equivalent to tabloids posting faked photos of some movie star's belly?

Re: Google contractors reportedly targeted homeless people for facial recognition

#108
post #48

Earlier quoted context omitted.

> I guess you're damned if you do, damned if you don't. This is not a case of that. Google (or its contractor) could easily have done this in a way that was not objectionable. They simply decided not to.

What would make it “not objectionable?”

For starters, telling the people what you're actually doing and maybe paying them more than 5 bucks.

Re: Google contractors reportedly targeted homeless people for facial recognition

#109
post #37
post #32

Earlier quoted context omitted.

I don’t think most people care about privacy all that much. Most live publicly with their faces on display for all to see and others taking it a step further, participating in Facebook alongside billions of others. It doesn’t scream facial identity being a major concern.

I think most people actually care quite a lot. That people live their lives accepting that their faces are on display is not evidence otherwise, since there is literally no other option. Participating Facebook is also not evidence otherwise -- at most, it's evidence that people are willing to trade privacy in some circumstances (and I think even that's a bit of a stretch), but I'll bet that most Facebook users would…

> it's evidence that people are willing to trade privacy in some circumstances

Which is what was being proposed and subsequently doubted: that people were willing to consensually trade their picture for $5.

Re: Google contractors reportedly targeted homeless people for facial recognition

#110
post #83

Earlier quoted context omitted.

Everyone who has uploaded a picture of themselves to Facebook, Linkedin or similar has already sold their face for free. That group includes almost everyone in the western world. $5 seems like a bargain compared to that.

to be clear, who has uploaded or had their photo uploaded; or potentially even attended any event someone bothered to setup a camera at.

No, people who had their photo uploaded by others didn't "sell" it. Selling implies an act (even if coerced).
Post reply on HN