Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms • Hacker News Board

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#141

Post by hoaw » Wed, Jan 23, 2019, 8:35 PM UTC

dgzl wrote: post #79

jpmcglone wrote: post #59

It’s more clear to me than ever now that the left and the right do not use the word “racist” the same way. The left thinks of racism in terms of outcome and the right thinks of racism in terms of intent. We could benefit from better language around these concepts, and honest dialogue about them too.

> We could benefit from better language around these concepts I believe this is a very under-rated problem today's public is facing. Not just these concepts, but many political terms are misused in today's climate, making for a conversationally ignorant population.

I don't see how it is being misused. This sort of distinction is pretty basic e.g. philosophy. People simply disagree. More and more people not understanding the basics in favor of whatever they read on social media might be a problem though.

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#142

Post by InitialLastName » Wed, Jan 23, 2019, 8:35 PM UTC

simion314 wrote: post #52

shadowbained wrote: post #30

Earlier quoted context omitted.

The problem here is, we don't even know what "fair" means. What do we mean by fair? Say we want to base fair on group outcomes, in which ways should we divide people into groups in order to perform some sort of balancing? Is hair color or texture a relevant way of subdividing people for instance? Are we being "fair" to red haired people? What if red haired people for some reason seem to be less likely to pay back loa…

Maybe don't split groups on genetics/biology/birth. Also make sure the data you use is not a proxy for the above criteria

But what if, due to historical happenstance, the data is already a proxy for groups split on genetics, biology, or birth? Using that data in a way that might create a feedback loop that propagates that happenstance is entirely likely.

Part of the problem is that, when dealing with humans as the data set, the data is really messy and there's lots of noise correlated to historical circumstance.

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#143

Post by eutropia » Wed, Jan 23, 2019, 8:35 PM UTC

saalweachter wrote: post #102

kerbalspacepro wrote: post #85

Earlier quoted context omitted.

Cults don't write about how to avoid being a cult, while the LW crowd has reams of texts and blog posts about how they are purposefully avoiding trying to become a cult.

Cults do talk quite a bit about why they're not a cult.

They even have an essay about _cultish countercultishness_: https://www.lesswrong.com/posts/gBma88LH3CLQsqyfS/cultish-co...

I think some of the feeling of cultishness might stem from a homogeneity of expression, but it kinda makes sense for a community of people dedicated to being rational -- after all, two rational agents with identical priors cannot "agree to disagree" in the mathematical sense. The community would tend to converge on common "correct" things and would have similar rebuttals for "incorrect" things, right?

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#144

Post by macspoofing » Wed, Jan 23, 2019, 8:35 PM UTC

SketchySeaBeast wrote: post #49

macspoofing wrote: post #43

Earlier quoted context omitted.

If AOC is an example of this new generation, we're in trouble.

In terms of tech literacy, which was my point, give this example I think she seems like she's doing quite well, regardless of her personal politics.

What tech literacy? If by 'tech literacy' you mean 'social-media literacy', or 'millennial-culture literacy', then I agree. Otherwise, I've never heard anything from AOC that demonstrates any kind of tech literacy.

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#145

Post by Obi_Juan_Kenobi » Wed, Jan 23, 2019, 8:36 PM UTC

darkpuma wrote: post #40

ThrustVectoring wrote: post #28

The second half of the article that talks about Yudkowsky, rationalists, and Roko's Basilisk is absolute trash and a gross mischaracterization of the actual positions taken. >Yudkowski has for more than a decade pursued the possibility of perfect human reasoning His website is literally called "Less Wrong". It's fundamentally a quest for improvement , not perfection. >His system of coldly logical reason, it turned ou…

> "This is because of the regretful (but understandable) decision to censor discussion of it on Less Wrong, which backfired spectacularly." Can you explain the (understandable) rationale behind Less Wrong censoring discussion of Roko's Basilisk? I've been lead to believe it was censored out of fear that discussing it could somehow inspire its actual creation. Honestly from a distance Less Wrong has always struck me a…

I don't fuck with LW, but SCC definitely has that culty vibe.

Overall I think Scott has some interesting articles and the general idea of rationalism is .. fine. But that's not at all what the communities around it are like. You just get tons of 'rational racism' with HBD (human biodiversity for those not in the know, which is basically just race realism with a nerdy veneer), IQ worship, and a group that collectively abandons anything like the rationalist ideal and becomes a post-hoc clusterfuck.

The community can almost be thought of like these racist AIs: garbage in, garbage out. They cling to any bit of rational-sounding data that supports their preconceptions, and have to adopt severe anti-academic stances to reject the mountains of data that contradict their views.

I suppose the deeper rot is that these rationalists fancy themselves like scientists, able to interact with primary data and draw conclusions, but any actual expert (i.e. someone that has performed research and published it in a community of peers, not whacko journals) is just confused how anyone could be so wrong. You regularly see posts where someone comes in who actually knows what they're talking about, authoritatively shows that all the nonsense they've been talking about makes no sense, and at best they get a 'huh, interesting' before the nonsense picks back up.

The idea of a rationalist community is to organize around questioning your own assumptions, confronting diverse views, and engaging in good-faith debate. But the reality is that basically none of that happens for anything controversial to the community, so it's just a veneer of legitimacy for in-group circle-jerking. And it just so happens that the in-group tends to be pretty damn racist.

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#146

Post by tomp » Wed, Jan 23, 2019, 8:36 PM UTC

kauffj wrote: post #11

There is no design of AI that can be simultaneously utilitarian, procedurally fair, and representatively fair. I'm going to repeat this again, since many people struggle with this. It is _literally impossible_ to achieve the best utilitarian outcome, the most procedurally fair outcome, and a representatively fair outcome. Anyone designing algorithms will have to make tradeoffs along this frontier. For a far better ar…

More to the point, “procedural fairness” or “equality of opportunity” (if I understand it correctly) is directly opposed to “representative fairness” or “equality of outcome”. In many situations, prioritising one will compromise the other. As an example, Harvard admission of Asian students.

Perspectives differ on which one is more important. Personally, I’m absolutely for the former, but it increasingly seems I’m a part of the minority (or the silent majority).

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#147

Post by squozzer » Wed, Jan 23, 2019, 8:36 PM UTC

Where I think the non-racists have failed us (meaning those whose attention lies mostly elsewhere) is in providing a description of a non-racist system. It may prove impossible, and I think if someone had described one, we'd all have heard about it.

Right now, the only system I can imagine as non-racist is closed and all of its elements belong to the same race.

Mathematicians would probably call this a trivial description. Comedians might call it a country club.

That moves the question not to whether we can build a system that isn't racist, but what racist qualities need the most suppression and which ones (where we have to choose) can be ignored.

Keep in mind this discussion has been ongoing for centuries and was simpler (e.g. slave vs. free) in the past than today (health, education, economics, psychology, religion.)

And more variables appear regularly.

That said, could algos help judges make more consistent rulings/sentences? Probably already are.

Judge: I'm gonna throw the book at this defendant!

LAL 9001: I can't let you do that.

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#148

Post by hannasanarion » Wed, Jan 23, 2019, 8:37 PM UTC

perl4ever wrote: post #108

paulmd wrote: post #47

Earlier quoted context omitted.

> Is hair color or texture a relevant way of subdividing people for instance? Are we being "fair" to red haired people? What if red haired people for some reason seem to be less likely to pay back loans? Machine learning is going to be very good at identifying proxies for features we don't like, even if we don't directly provide those features. Weighting red hair color could be a proxy for the Irish, targeting hair t…

"And we have to find a way to make sure that The Algorithm (tm) doesn't just find a way to machine-learn redlining or some other proxy for race or other protected characteristics." This is not wrong, but it boils down to "don't make mistakes when optimizing functions". It's offensive because it's coming from someone who has nothing to contribute except "do your job better". It's like saying we have a problem with pla…

The consequences of a poorly designed algorithm are far less visible than those of a poorly designed airplane. Oftentimes they are deliberately kept secret. How many banks do you know that publish their loan qualification source code?

Mechanical engineers are taught from the beginning of their careers to think carefully about the consequences of their decisions.

Computer engineers, not so much. We're taught how to optimize, and make machines that optimize, and make optimizers that optimize our optimizing machines, but it's pretty rare to encounter a data ethics course, and they're almost never program requirements.

Facebook got in trouble for algorithmic redlining in 2013, serving loan sharky ads to black people (which its algorithms learned via proxy features, like zip code). Does Facebook hire third-rate engineers who would point the engine on a plane the wrong way? No! Facebook has some of the best people in the industry! But even the best people in the industry were and in many cases still are not fully considering the consequences of our designs, because we weren't trained to.

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#149

Post by cma » Wed, Jan 23, 2019, 8:37 PM UTC

chongli wrote: post #56

darkpuma wrote: post #40

Earlier quoted context omitted.

> "This is because of the regretful (but understandable) decision to censor discussion of it on Less Wrong, which backfired spectacularly." Can you explain the (understandable) rationale behind Less Wrong censoring discussion of Roko's Basilisk? I've been lead to believe it was censored out of fear that discussing it could somehow inspire its actual creation. Honestly from a distance Less Wrong has always struck me a…

I think the more charitable interpretation is that the discussion was taking over everything. If people are fixating on a narrow topic for a long time (without much progress) then sometimes it's helpful to say "alright people, let's move along now." You're right about Less Wrong feeling like a cult, though. It gives me the heebie-jeebies.

>You're right about Less Wrong feeling like a cult, though. It gives me the heebie-jeebies.

They even have a succession plan similar to finding the next Dalai Lama:

https://www.youtube.com/watch?v=d9_2qhkWHN8

Re: Alexandria Ocasio-Cortez Is Absolutely Right About Racist Algorithms

#150

Post by macspoofing » Wed, Jan 23, 2019, 8:37 PM UTC

akhilcacharya wrote: post #51

macspoofing wrote: post #41

There's a tendency in the media to 'clean up' the silly things that Alexandria Ocasio-Cortez consistently says, to make them sound half rational.

Except..she's right...

About what? She is a far left progressive with a very shallow understanding of issues (she literally just learned what a marginal tax rate is). I understand you may share her political beliefs, shallow that they may be, but that doesn't make her 'right'.