Thursday, August 17, 2017

The Obverse Godwin's Law

This meme recently showed up on CNN host Chris Cuomo's Twitter feed (and elsewhere):
Pretending that people who insist on punching Nazis as private citizens are the moral or legal equivalent of the men who stormed Normandy is not only false, it is idiotic and ahistorical. Allow me to introduce you to Mssrs. Harold Sturtevant and E.C. Lackey, USN:

Harold Sturtevant was a sailor in the United States Navy. In January 1941, he and fellow sailor E.C. Lackey climbed up the fire escape of the building which housed the German consulate in San Francisco, California and slashed and tore down the flag of Nazi Germany which was flying there in honor of the 7th anniversary of the founding of the Third Reich. The two men were arrested, tried, court martialed for malicious mischief and received a dishonorable discharge from the Navy. The German Foreign Ministry protested the incident and the United States Department of State expressed their regrets.
 Which is to say, the United States, then being at peace with the German Reich, issued a formal apology and punished the perpetrators. Yes, the modern white supremacists are odious and worthy of full-throated denunciation (something President Trump apparently couldn't quite bring himself to do), but punching Nazis doesn't make you a hero, it just makes you a thug and a vigilante. Mike Godwin's famous law about discussions eventually hauling in Hitler and Nazis is fine as far as it goes; his point was to force people to use such analogies thoughtfully. But it has no obverse, and we desperately need one.

Monday, August 14, 2017

Vox Ladysplains The #GoogleMemo, And Other Related Stuff

I want to start with Cynthia Lee's "ladysplain"-ing of the #GoogleMemo in Vox, not because it is good but because it serves as an exemplar of how the blank slatists insist on misreading James Damore's essay on the futility of Google's approach to coder "diversity". She opens by writing that "It’s important to appreciate the background of endless skepticism that every woman in tech faces, and the resulting exhaustion we feel as the legitimacy of our presence is constantly questioned." This recalls work by Roy Baumeister in which he observes that putdowns are endemic to male culture, a constant reminder that respect is earned and in limited supply:



Lee's reaction to this shows exactly how right Baumeister was when he wrote
This, incidentally, has probably been a major source of friction as women have moved into the workplace, and organizations have had to shift toward policies that everyone is entitled to respect. The men hadn’t originally built them to respect everybody.
 Her next complaint is against the "sleight-of-hand" of averages she claims Damore uses that she claims turns women "against their own gender." However, Damore is very careful to note that, "Many of these differences are small and there’s significant overlap between men and women, so you can’t say anything about an individual given these population level distributions. This doesn't matter; Lee has no interest in the real, measured preferences of populations, and as we see in the next section, this has catastrophic consequences for her argument (emboldening mine):
If, as the manifesto’s defenders claim, the population averages do not have anything to say about individual Googlers, who are all exceptional, then why is Google the subject of the manifesto’s arguments at all? What do averages have to do with hiring practices at a company that famously hires fewer than one percent of applicants? In the name of the rational empiricism and quantitative rigor that the manifesto holds so dear, shouldn’t we insist that it only cite studies that specifically speak to the tails of the distribution — to the actual pool of women Google draws from?
Funny you should ask. That, actually, is the exact problem, and the fact that Lee misses it is unsurprising. Implicit in her argument here is the idea that men and women, taken as populations, will be interested in exactly the same thing, so that by the time you get to that narrow tail, you will have exactly the same number of individuals. This is categorically false; Damore cited evidence that, on average, women have more interest in working with people rather than things. The narrow tail of people interested in thing-work is where Google is hiring. She addresses this aspect of the population not at all. This is the crucial part of his argument, and indeed is what we see in practice, as her very next example demonstrates!
For example, we could look to the percentage of women majoring in computer science at highly selective colleges and universities. Women currently make up about 30 percent of the computer science majors at Stanford University, one key source of Google’s elite workforce. Harvey Mudd College, another elite program, has seen its numbers grow steadily for many years, and is currently at about 50 percent women in their computer science department.
Yet as Scott Alexander showed, MIT and Harvey Mudd get their female graduation rates by stuffing the pipeline with more women than most institutions that don't discriminate:
...MIT admitting 2x more women than men matches nicely with their computer science department being 40% women (= 2x the national average of 20%). Harvey Mudd admitted 2.5x more women than men matches nicely with their computer science being 55% women (just a hair over 2.5x the national average of 20%). Plus everyone in this discussion agrees that a bunch of colleges are desperately trying to admit as many women as they can to get even close to parity in CS.t
(While I don't have figures for Stanford, it would certainly be interesting to learn the percentage of female applicants accepted into their program. I would not be shocked at all if they did the same as Harvey Mudd, and indeed recent figures make it appear that is the case.) The reality is that computer science, and engineering more broadly, has been stuck at about 20% female (or less) for decades, regardless of gimmicks. Using two cherry-picked institutions that in turn cherry-pick their candidates is a perfect example of Damore's argument: they haven't magically found a way to get girls to like coding so much as they've found more girls who do (at the expense of other institutions' admissions). The strong argument would be explaining away why the female CS/engineering population is what it is at, say, Iowa State, or at a random sample of universities. Lee does not attempt it. (I also note in passing she does not wrestle with something Alexander observed, and that is that CS/engineering gender parity is best in nations such as Zimbabwe and Thailand, countries "not exactly famous for [their] deep commitment to gender equality.") In so doing, she cedes her entire argument.



Other linky goodness before (hopefully) closing this chapter:
  • David Brooks thinks Sundar Pichai should resign as Google CEO.
  • Conservatives are lining up to protest Damore's firing, via The Hill:
    Right-wing activist Jack Posobiec and a coalition of free speech groups are organizing marches against Google next week to protest Damore’s firing.

    “We are going to raise awareness about Google’s one-sided bias and campaign against dissenting opinions and voices,” Posobiec told The Mercury News on Thursday.
  • Also at The Federalist Bre Payton finds the media broadly insists on misreading Damore's memo, as one would expect.
  • Robert Tracinski writes in The Federalist about "the Google inquisition":
    A Wired profile digging into Damore’s personal history (this is politics now, so we do opposition research) gives us this description: “Damore’s fellow students at Harvard remember him as very smart but awkward around people.” Gosh, it would be a real shame if people like that were allowed to be hired in Silicon Valley.

    I’m joking, of course, because this is precisely the kind of personality that built Silicon Valley. But maybe not any more. Yet that’s not the biggest, most dangerous part of this story for Google and the other tech giants. The most dangerous part is that they are now beginning to be seen by the public (or revealed, depending on how you look at it) as politicized entities. Politicized entities to whom we are giving enormous amounts of data on our lives, thoughts, and interests.
    Surveying this landscape, it's easy to imagine how the politics will align going forward. Having collected Damore's scalp on the basis of merely acknowledging that men and women have divergent interests (as populations), what else could the diversity mavens accomplish with sufficient dudgeon behind them? I have previously noted Anita Sarkeesian's censorious tendencies. It does not seem even a slight stretch, given the Euros have introduced a "right to be forgotten", that Google is or will soon be on the list of targets. That is, at some future date, they will demand wrongthink such as Damore's be banished to PageRank purgatory. No pinnacle in modern content distribution is higher, and for that reason we must fight this at all costs.
Update 2017-08-15:
Damore got an op-ed into the Wall Street Journal defending himself. Excerpt:

Everything changed when the document went viral within the company and the wider tech world. Those most zealously committed to the diversity creed—that all differences in outcome are due to differential treatment and all people are inherently the same—could not let this public offense go unpunished. They sent angry emails to Google’s human-resources department and everyone up my management chain, demanding censorship, retaliation and atonement.

Upper management tried to placate this surge of outrage by shaming me and misrepresenting my document, but they couldn’t really do otherwise: The mob would have set upon anyone who openly agreed with me or even tolerated my views. When the whole episode finally became a giant media controversy, thanks to external leaks, Google had to solve the problem caused by my supposedly sexist, anti-diversity manifesto, and the whole company came under heated and sometimes threatening scrutiny.

Monday, August 7, 2017

The Google Heretic

The news arrived a couple days ago that Google's engineering ranks include someone with unorthodox opinions on the subject of "diversity" as practiced in Silicon Valley, the actual text (minus graphs and hyperlinks) leaking out in the pages of Gizmodo Saturday. The essay itself senselessly adopts some of the worst flaws of modern political discourse from the left, particularly "psychological safety", a fatally damaged concept that has no place in grownup discussion. He (I assume the author is male) also mislabels as "authoritarian" the idiotic and badly misguided private efforts toward an inflexible and unachievable "diversity" goal; I note the author is free to leave Google, and work elsewhere. He also adopts the whiny language, itself extracted from Marxism's leaden skeins, that makes so much feminist writing unbearable: what does "swaths of men without support" even mean?

But those criticisms aside, the author is right about biological origins of a number of disparities between the sexes, particularly in mathematics, which are of long-standing and universal at the higher end of achievement. That is to say, from a population standpoint, women are more uniform in ability than men, and thus you end up with fewer geniuses — and fewer morons. (There are nations where female averages are actually higher than male averages [PDF, see page 10], but male-female average math score gaps exist for the majority of OECD countries save Iceland, where it is reversed, with some less significant than others.) Unfortunately, he does not provide substantiation for his claims, unless of course Gizmodo's editorial decision to strip the jeremiad of hyperlinks was an act of deep political cowardice.

However, why did he feel it necessary to pen such a document? To know that, it is necessary to ask, how is it that Google has a Vice President of Diversity, whose job presumably is to root out and destroy a would-be modern T.J. Rodgers accidentally joining the Googleplex ranks? Google, simply, has become a huge visible success, to the extent that it can afford to operate many companies with dubious or nonexistent paths to profitability. Throwing some bones to the commercial feminists is a no-brainer, for now; if you can lose a billion dollars in a quarter, you're doing something right. But as with Microsoft and its seemingly invincible computing platform that took a dive once they made (wholly necessary) forays into mobile, nothing is guaranteed, and today's juggernaut could easily be tomorrow's roadkill. Advanced parasitism of that kind will have no place in a smaller company, either devourer or devoured. Asking "why are there so few women coders?" is as pointless as asking "Why are there so many Jews in Hollywood/banking/diamond cutting?" The answer is, and should always be, who cares?

Update, 2017-08-08: Google yesterday fired author James Damore on the ground that he was  "perpetuating gender stereotypes", thus essentially proving he was right about the company acting as an echo chamber. (I would observe that companies set up thusly are also liable to fall apart in other ways.) As usual, Scott Alexander has a terrific followup:
Galpin investigated the percent of women in computer classes all around the world. Her number of 26% for the US is slightly higher than I usually hear, probably because it’s older (the percent women in computing has actually gone down over time!). The least sexist countries I can think of – Sweden, New Zealand, Canada, etc – all have somewhere around the same number (30%, 20%, and 24%, respectively). The most sexist countries do extremely well on this metric! The highest numbers on the chart are all from non-Western, non-First-World countries that do middling-to-poor on the Gender Development Index: Thailand with 55%, Guyana with 54%, Malaysia with 51%, Iran with 41%, Zimbabwe with 41%, and Mexico with 39%. Needless to say, Zimbabwe is not exactly famous for its deep commitment to gender equality.
As usual, the whole damned thing is worth reading.Also, this:






And, this:
Update 2017-08-08 9:31: Also a good read at In A Crowded Theater:
Google CEO Sundar Pichai acknowledged that these topics are “fair to debate.”   He claimed that Googlers are free to discuss these topics so long as they do not advance “harmful gender stereotypes in our workplace.”  But by firing Damore, Pichai belied any commitment to real discussion.  A true debate about these issues requires grappling with all of the thorny premises.
 The good news is that now a woman can be hired to replace Mr. Damore, thus ensuring Google's comittment to diversity. Also, they will make sure she has the right opinions before hiring, and the Vice President of Truth Diversity to make sure she keeps them with the party line.

Update 2017-08-08 9:58: Also Inez Feltcher at The Federalist:
Damore is guilty of nothing more than gently stating the obvious truth, backed by a laundry list of scientific studies: on average, men and women have divergent talents, interests, and skills. Because of these differences, men and women make different career decisions in the aggregate. Damore’s great offense was recognizing that maybe, just maybe, the imbalance between men and women in software engineering has more to do with freedom of choice than being the six-figure salary counterparts to the handmaids in Gilead.

Instead of fighting these “gaps” as the result of discriminatory systems and attempting to force men and women to be the same, we should consider the possibility that their divergent choices are the result of our true diversity.
Also useful in that Federalist piece is a link to the essay, links intact. As I expected above, the stripped links point at buttressing evidence for his thesis, which Gizmodo made the political decision to shamefully omit in their rebroadcasting.

Update 2017-08-09: Before passing on this subject, it's worth quoting this passage from the Scott Alexander post upthread:
We know that interests are highly malleable. Female students become significantly more interested in science careers after having a teacher who discusses the problem of underrepresentation. And at Harvey Mudd College, computer science majors were around 10% women a decade ago. Today they’re 55%.
I highly recommend Freddie deBoer’s Why Selection Bias Is The Most Powerful Force In Education. If an educational program shows amazing results, and there’s any possible way it’s selection bias – then it’s selection bias.

I looked into Harvey Mudd’s STEM admission numbers, and, sure enough, they admit women at 2.5x the rate as men. So, yeah, it’s selection bias.

I don’t blame them. All they have to do is cultivate a reputation as a place to go if you’re a woman interested in computer science, attract lots of female CS applicants, then make sure to admit all the CS-interested female applicants they get. In exchange, they get constant glowing praise from every newspaper in the country (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, etc, etc, etc).
Now, we don't know if female CS candidates are admitted at 2.5 times vs. men — the rate could be higher or lower in that specialty — but this definitely points to at least a potential problem for their headline story about women in CS: it's not so much that they support women as they throw enough women at the problem that eventually some of them will get through.

Look! More stupid!




Tuesday, August 1, 2017

I'm Entitled To Your Opinion Dep't: Mehera Bonner Reviews Dunkirk

The meat of her criticism is actually apt; the film doesn't appear to have an actual plot, but instead is a series of pastiches of interwoven stories: the older man (Mr. Dawson) piloting a private pleasure boat to rescue Tommies on the Dunkirk beach, the outnumbered fighter pilots taking on relentless Luftwaffe adversaries hectoring ground troops and sinking transports, the young soldier separated from his unit trying against the odds to make it home (Tommy), the commanders in charge of moving the men off the beach and onto the absurdly small (and shrinking daily) numbers of available military transports. That is, she's not wrong in this specific complaint.

But the film itself has earned a great deal of praise, and deservedly so, despite the overall failing of lacking apparent narrative. Part of that is because we already know the outcome: Britain's fathers came to the rescue of her sons, and even a significant number of French troops as well. It is beautifully photographed, flawlessly acted, and rippling with dramatic tension from the opening until almost the close. None of these virtues apparently appeal to Bonner:
But my main issue with Dunkirk is that it's so clearly designed for men to man-out over. And look, it's not like I need every movie to have "strong female leads." Wonder Woman can probably tide me over for at least a year, and I understand that this war was dominated by brave male soldiers. I get that. But the packaging of the film, the general vibe, and the tenor of the people applauding it just screams "men-only"—and specifically seems to cater to a certain type of very pretentious man who would love nothing more than to explain to me why I'm wrong about not liking it. If this movie were a dating profile pic, it would be a swole guy at the gym who also goes to Harvard. If it was a drink it would be Stumptown coffee. If it was one of your friends, it would be the one who starts his sentences with "I get what you're saying, but..."
How terrible — someone makes movies that appeal to men? Her reaction isn't quite "THIS MUST STOP NOW", but you can hear her mentally outfitting anyone who actually likes the film with an invisible fedora (the universal headgear of the MRA). The idea that men died in battle so that someone like Bonner could spout narcissistic and childish opinions is itself cringe-worthy, but as Kyle Smith ably answers in National Review Online, the problem is really a branch of the Annie Wilkes model of culture (emboldening mine):
In a moment of clarity I understood what the two main imperatives of higher education were to Absurd Feminist and to so many of her peers: First, instead of broadening her horizons and taking her outside herself to discover the world, she demanded the educators filter all knowledge through her own experience to make it relatable to her. Second, all learning was to be valued in proportion to how effectively it could be made into a cudgel in the identity-politics war. Dispatches, with its virtually all-male cast, represented a pernicious advance for the patriarchy, even if it was about the agonies suffered by men.
It seems unlikely that Marie Claire’s reviewer, Mehera Bonner, has before her an exceptionally bright career of writing about film. As for a career of writing about feminism, though, the sky, for Bonner, is the limit. Her essay could plausibly have appeared on any number of bristling feminist sites. What is her reasoning except feminism taken to its logical extreme? Feminists often declare to the world that they stand merely for an entirely reasonable proposition — say, that women’s lives are as important as men’s. Who would dispute that? Yet feminist writing usually continues far past this point into a need to prove women and men have been equally important in every context, even in history. If women turn out to be mostly irrelevant to an incident, then it is the moral duty of socially conscious creative artists to ignore the matter. They should retrain their sights on something that will give absurd feminists something they can relate to, something that will advance the cause of feminism in general.
 She doesn't like the movie; fine, we get that. But as Smith observes, "Feminism means constant maintenance of an imaginary set of scales, and she fears Dunkirk adds weight to the masculine side, tipping the culture away from women." What could be more absurd?