Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Attacking discrimination with smarter machine learning

research.google.com

Post reply on HN
Print view View original

121–130 of 201 posts

Re: Attacking discrimination with smarter machine learning

#121

Post by Retric »

striking wrote: post #111
Retric wrote: post #78

Earlier quoted context omitted.

Now, this is his first offence, but a young man is found high on PCB walking down the street hitting cars with a baseball bat. It takes five cops and a significant struggle to arrest him. Having read that, what would you assume is race and economic background is? After criminal proceedings he was sentenced to community service. Now, what would you assume is race and economic background is? PS: Bias is insidious and r…

I'd assume he's a poor white man, from the very beginning. White, because most people in the US are white.

Interesting you said poor, at the time he had over 6 million in a trust fund and stood to inherent vastly more. Some people change their minds because of the community service vs. prison but some don't.

PS: PCP has a reputation as a rural poor white persons drug. I wonder if you would have had a different impression if I started by saying he got community service.

Re: Attacking discrimination with smarter machine learning

#122

Post by tptacek »

BickNowstrom wrote: post #82
tptacek wrote: post #24

Earlier quoted context omitted.

The difference with machine learning is that the model isn't designed by humans, through an actuarial process we can keep our brains wrapped around. It's a black box. We are OK attributing "crash risk" to young male drivers, because we can observe both that they are as a cohort statistically likely to crash and also understand why that would be the case. On the other hand, we're not comfortable with the idea that a c…

It is possible to inspect black box models. See for instance http://www.blackboxworkshop.org/pdf/Turner2015_MES.pdf and https://homes.cs.washington.edu/~marcotcr/blog/lime/ . It is also common to use highly accurate white box models for cases where it is important to not overfit, leak, or discriminate, like MARS and GA^2M (somewhat out of scope for an actuary using R). If a computer program spots an irrelevant unhelp…

Isn't that the kind of work we're commenting on with this story?

Re: Attacking discrimination with smarter machine learning

#123

Post by tjic »

KirinDave wrote: post #92
tjic wrote: post #29

Earlier quoted context omitted.

Please don't say "we". Not everyone shares your politics. I'm perfectly happy with algorithms detecting that certain people are more likely to be safe drivers than average, and giving them lower rates, and concentrating premiums on the groups more likely to be in accidents, even if I don't understand why Armenians (in your example) get in more crashes.

I've run into many people who share this sentiment, and it always surprises me. I've never once met a person who was cheerful about experiencing algorithm-driven prevenge. I think it's very easy to sit back and say "I think we should let this happen!" if they've never knowingly experienced loss based on this phenomenon. A great example is how very resentful many young white men of college age are that universities ar…

> . I think it's very easy to sit back and say "I think we should let this happen!" if they've never knowingly experienced loss based on this phenomenon.

I'm sure that I've experienced increased costs based on this.

The issue is that I don't consider that the morality of forcing my will on other people depends at all on whether their current behavior is advantageous or disadvantageous to me.

Re: Attacking discrimination with smarter machine learning

#124

Post by tjic »

tptacek wrote: post #34
tjic wrote: post #29

Earlier quoted context omitted.

Please don't say "we". Not everyone shares your politics. I'm perfectly happy with algorithms detecting that certain people are more likely to be safe drivers than average, and giving them lower rates, and concentrating premiums on the groups more likely to be in accidents, even if I don't understand why Armenians (in your example) get in more crashes.

You've just defined away the problem. It's not that we're not OK with computers accounting for the idea that Armenians get into more car accidents (we might or might not be). It's that we don't know if that's actually the case, because ML-generated models aren't that simple (if they were, we wouldn't need ML, just a bunch of actuaries using R). The distinction is that with young male drivers, we have two supporting c…

> You've just defined away the problem.

What's the null hypothesis here? From my point of view, you're the one who defined into existence as a problem something that is not a problem.

Re: Attacking discrimination with smarter machine learning

#125

Post by tjic »

tptacek wrote: post #120
chronic6l wrote: post #113

Blacks (or African Americans) are less intelligent than other ethnicities, in average. We have information that satisfies your two criteria of supporting class information. Yet people on HN (liberals/progressives) will get upset with this fact. Supporting classes: 1. Through employment, high school graduation, incarceration, and homicide rates. 2. Due to black cultural, male macho-independence, discrimination from no…

"Blacks" are not "less intelligent" than "other ethnicities". Please race troll somewhere else.

chronic6l has a wealth of peer-reviewed science on his side, tptacek.

While I imagine that most of us think it would be nicer if all subgroups had the same mean attributes, that's not the world we live in and we shouldn't turn our backs on science and embrace faith-based arguments, no matter how nice the motivations.

Re: Attacking discrimination with smarter machine learning

#126

Post by jtbigwoo »

Bartweiss wrote: post #72
corey_moncure wrote: post #61

Earlier quoted context omitted.

> The data tells you that a black person is more likely to be a criminal than a white person. There are two possible reasons for this: I'd like to add a third possible reason for your consideration. Since "criminality" i.e. guilt of committing a crime is determined after a process engaging the law enforcement and justice systems, we have to examine whether there are inherent biases in those systems that result in ske…

This is definitely a point worth raising. I'll defend the parent comment by observing that we can reproduce this effect with non-judicial metrics, like "murder rate in majority foo-race communities". Assuming the reporting rate for murders is very high across the board, this escapes bias in both policing and conviction rates, and sends us back to other explanations. Even so, the policing/judicial question is really i…

>> I'll defend the parent comment by observing that we can reproduce this effect with non-judicial metrics, like "murder rate in majority foo-race communities". Assuming the reporting rate for murders is very high across the board, this escapes bias in both policing and conviction rates, and sends us back to other explanations.

A community has a fair amount of power in deciding where crime happens. The police (by policy or culture) can choose to push illegal activity to certain neighborhoods. If the police come down hard on drugs, prostitution, etc. in white neighborhoods, then that activity will move to non-white neighborhoods. The problems that come along with increased illegal activity, including higher murder rates, will be concentrated in non-white neighborhoods.

Re: Attacking discrimination with smarter machine learning

#127

Post by tptacek »

tjic wrote: post #125
tptacek wrote: post #120

Earlier quoted context omitted.

"Blacks" are not "less intelligent" than "other ethnicities". Please race troll somewhere else.

chronic6l has a wealth of peer-reviewed science on his side, tptacek. While I imagine that most of us think it would be nicer if all subgroups had the same mean attributes, that's not the world we live in and we shouldn't turn our backs on science and embrace faith-based arguments, no matter how nice the motivations.

No, he doesn't. He has a mixture of early Jensenist psychometric research that has been superseded, and neo-phrenologists like Rushton that have been discredited.

But of course it's easy to drop little bombs like this into threads and put the onus on other people to explain the science. That's what makes it trolling.

Whatever snappy response you have for this, please spare us. There is a reason "the subject of racial IQ gaps" is a controversial scientific debate, and I'm quite confident you haven't settled it.

Also: please lose your condescending tone. There are something like 5 comments on this thread in which I attempt to explain how the problems this CS research (which you appear to find intolerable) are going to cause problems for conservatives, rural citizens, and libertarians. All I see from you is derisive comments and an inexplicable defense of an off-topic introduction of racial IQ gaps to the thread.

If you have something to say about computer science research, go for it. Please do not pretend that you're occupying some kind of rationalist high ground while militantly preventing other people from discussing computer science on HN. The person in this conversation most diligently trying to turn it towards politics is you.

Re: Attacking discrimination with smarter machine learning

#129

Post by striking »

Retric wrote: post #121
striking wrote: post #111

Earlier quoted context omitted.

I'd assume he's a poor white man, from the very beginning. White, because most people in the US are white.

Interesting you said poor, at the time he had over 6 million in a trust fund and stood to inherent vastly more. Some people change their minds because of the community service vs. prison but some don't. PS: PCP has a reputation as a rural poor white persons drug. I wonder if you would have had a different impression if I started by saying he got community service.

Look, if we're talking about assumptions, we're talking about probability. I have no idea who you're talking about. I'll happen upon some statistics once in a while, but otherwise will avoid making assumptions because I simply don't know. You asked me for what I'd say was most statistically probable, you got it.

I assume you mean "PCP" and not what circuit boards are printed on, yeah. You're right, I've never heard of a rich person taking PCP (cocaine, fancy liquor, etc. for them, right?) although I'm absolutely unsure of how PCP use breaks down racially. (If you have any source for those statistics, you've piqued my interest in them and I'd love a link)

Again, my guess of "white" was based on the fact that most people are white, and thus most cases of violence (without further stratification) are probably by whites. And what, do African-Americans not get community service? That probably wouldn't have changed my mind, and actually sounds a little bit racist.

Re: Attacking discrimination with smarter machine learning

#130

Post by tptacek »

tjic wrote: post #124
tptacek wrote: post #34

Earlier quoted context omitted.

You've just defined away the problem. It's not that we're not OK with computers accounting for the idea that Armenians get into more car accidents (we might or might not be). It's that we don't know if that's actually the case, because ML-generated models aren't that simple (if they were, we wouldn't need ML, just a bunch of actuaries using R). The distinction is that with young male drivers, we have two supporting c…

> You've just defined away the problem. What's the null hypothesis here? From my point of view, you're the one who defined into existence as a problem something that is not a problem.

What are you talking about? Read the article. It's about computer science, not your political sensitivities.
Post reply on HN