Live data from Hacker News

The Kolmogorov option

scottaaronson.com

471–476 of 476 posts

Re: The Kolmogorov option

#471
post #457

Earlier quoted context omitted.

The reality is that 2,000 of them are indistinguishably good and diversity programs are intended to make sure that google hires equally able women out of that group instead of choosing all the men.

Imagine a theoretical scenario where they have 500 female applicants and 9,500 male applicants, but desperately want a 50/50 workforce. Then they logically would hire all female applicants, including those who are not part of the 2000 most capable applicants. Using your number, with 2k out of 10k applications being indistinguishable elite, that means 1 in 5 applicants are suitable to be hired on merit. Assuming that…

You clearly have no idea how hiring works at major tech companies like Google.

1: Even if there are a 1000 positions, and only 1000 candidates, if they are underqualified, they do not get hired. Google, Amazon, Facebook, etc, try to have an objective bar, not one relative to the available candidates.

2: Nobody is pressuring for a 50/50 workforce. It's an ideal, and a reasonable-sounding long term goal, but even the most aggressive pro-diversity initiatives do not set a target of 50/50 ratio. In fact, in most cases HIRE targets aren't set at all. The targets are set for opportunities - ie number of diversity candidates evaluated or interviewed. The hiring process remains pure.

Source: Work for major top-10 tech company and do a ton of hiring, and diversity training.

Re: The Kolmogorov option

#472
post #34

Tell me, truly, is this a view held by many here? That, as the church denounced Galileo's observation that we turn around the sun as heretical, so too our modern PC culture suppresses legitimate inquiry into wether gender is a determinant of programming aptitude? This is a comparison worth drawing? Galileo was a scientist. He published papers, books, treatises. He devoted his life to the pursuit of the truth, found a…

> until equality is the rule of the day But people aren't clones and we aren't all equal. Hey, I want to be equal to LeBron James. Please, give me the same salary. You want equality, there's communism for you. See how that panned out. > women are innately unsuited to computers Did you read the original paper? Damore never wrote that. He was talking about statistical chance, not generalizing like you are now.

It's not equality of results, it's equality of opportunity.

You can get the same initial chance to prove your ability at basketball as he did. You can even get a slight boost to compensate for things (maybe he was a lot more motivated, or maybe he got some practice in the 'hood). And it'll turn out you have no skills/talent/ability, you're not 2 meters high, and so on.

Now, with sports, it's not a problem if we don't spend more than a few seconds of thought on trying to create equal opportunities, because with competitive sports the point is exactly to not have equal outcomes.

But with broad categories like jobs/career, income, wealth, political/legal/professional representation, we pretty much want to make sure that if one gender is underrepresented then it's not hurting us all on the long run. It might turn out that yeah, women don't really like tech, and those few who do are very welcome, men can behave around them, and those instances of sexism are just the expected rare problems that do happen in every other walks of life. But so far, it seems that it's very much not the case.

Re: The Kolmogorov option

#473
post #54

Earlier quoted context omitted.

I think you missed the point of that one. He doesn't want to hurt other people, but realizes he won't know all the ways he could do so accidentally. So, in the interest of not hurting people, if he is unsure whether something will or not, he'll defer to smarter people about what's appropriate behavior. If you don't care whether what you say hurts others, you need do nothing, need defer to no one any time, ever.

What if you're deferring to smarter people who want to hurt others though? Quite the conundrum.

Well, if you were a common Hans in 1930 Germany, you might have listened to smarter folks, and they wanted to hurt Jews a lot, take their stuff and beat them for their greed (or whatever).

Are they really smarter? If you think yes, then what can we do, you'll happily join them, and probably won't be swayed by even smarter people's arguments.

If on the other hand those allegedly smarter people advocate against the spread of nationalism by recommending that you maybe don't employ or get employed by nationalists, that might be a bit different, yet maybe totally counterproductive.

Re: The Kolmogorov option

#474

I wrote the present comment to 1) teach something about mathematics and probability 2) share enlightenment. In summary this comment should change your thinking about all subjects, forever. -------------- Program of study for this comment: 1. Read section 1 (approx. 1 hour.) Goal: improve your mathematical reasoning. 2. 30 minute break. 3. Read section 2 (approx 5 minutes). Goal: enlightenment. 1. The most important m…

Appreciate you trying to share some insight. But from the hype, I was expecting some kind of world changing revelation. But basically you're just explaining the basics of bayesian statistics?

A lot of people already know this stuff. It's part of an intro course in probability/stats that you'd take while doing a CS degree, or a course about experiments for other science degrees.

Re: The Kolmogorov option

#475
post #429
post #420

Earlier quoted context omitted.

> he has done no research on the subject at all. Except for the citations. > why I could easily tell that the author of the manifesto is both ignorant and uninterested in the subject Continuing to attack the credibility of a largely factual statement (the memo in question) is still not compelling. If there was evidence supporting opposing viewpoints, even hypothetically, you might want to lead with that in your inter…

I am not attacking the cited papers, just their relevance. He might as well have said, "we should not fight global warming because of special relativity". Special relativity may be true, but if you think it relates to global warming, then you have done no research on global warming, as you clearly don't understand the problem. That he thinks that innate differences in interest between the sexes relates to the problem…

As long as you don't explain why you think it is not relevant, there'll not be much progress.

You could argue for example the default should be to treat the two(representation of women in tech vs psychological gender differences) different and "burden of proof" should be on the side arguing they're same. I think that's not obvious to me(and probably a lot of others).

Re: The Kolmogorov option

#476
post #429

Earlier quoted context omitted.

I am not attacking the cited papers, just their relevance. He might as well have said, "we should not fight global warming because of special relativity". Special relativity may be true, but if you think it relates to global warming, then you have done no research on global warming, as you clearly don't understand the problem. That he thinks that innate differences in interest between the sexes relates to the problem…

As long as you don't explain why you think it is not relevant, there'll not be much progress. You could argue for example the default should be to treat the two(representation of women in tech vs psychological gender differences) different and "burden of proof" should be on the side arguing they're same. I think that's not obvious to me(and probably a lot of others).

The simplest way I can think of putting it is as follows: Say there was a chemical plant dumping large amounts of radioactive waste into a stream, and the population nearby is experiencing an increased incidence rate of cancer. People call on the plant to stop polluting, and the owner cites some papers showing that cancer has other causes too. Those studies would be completely irrelevant, and a clear attempt by the plant to find an excuse to not stop polluting.

That the participation rate of women in software is experiencing a sharp decline in recent decades, mostly in the US, (while, say, women participation in physics is on the rise) is an undisputed fact; that women working in software companies report abuse is a similarly undisputed facts. That is the problem we're trying to solve. That there may be contributing immutable biological factors is completely irrelevant. The main difference between this and the chemical plant analogy is that while there is no doubt that there are causes of cancer other than exposure to radioactive waste, the studies the author of the manifesto cites are very much controversial on purely scientific grounds, plus he clearly does not understand their context (e.g. that women in liberal countries prefer to work less etc. is no indication whatsoever of any innate factors).

But the main problem is this: Damore is no expert on the subject. His belief that the studies he cites are relevant is based on a layman's intuition, which is wrong in this case, as in many others. Explaining exactly the problems is much harder than bringing up ridiculous hypothesis by people unfamiliar with the subject like Damore. If you want to know exactly what the issue is (and I doubt Damore does, because he clearly read no research on marginalization of women), you should read actual scholarship on the subject, and not rants by some random guy.

Post reply on HN