Sunday, February 12, 2017

Choosing-Your-Poison Isn't That Difficult

In last week's essay, I linked without comment to Jack Baruth's essay on that execrable Audi superbowl ad. Today, Trumwill points out that the Right also has its car commercials:

These two commercials actually serve pretty well as synecdoches for the political metaphysics of the two halves of our current political spectrum:

  • Right: celebrate our common American identity for the purpose of kicking other countries' butts in economic competition.

  • Left: celebrate our (your) common class identity for the purpose of kicking flyover country's butts in status competition.

Although it would be refreshing to find someone actually defend the Left's metaphysic instead of striking a pose of choose-your-poison neutrality.

Sunday, February 05, 2017

Sexbots and Flying Cars

I'm going to make a prediction (and you heard it here first):

Sexbots will be to the 21st Century what flying cars were to the 20th Century

It seems that a week can't go by this past year without me stumbling across an article like this one:

This got me thinking about flying cars.

Wikipedia dates the flying car concept to the early 20th Century, but it was certainly a futurology staple by the golden age of sci-fi. It's role in pop-culture today is mainly a synecdoche for our disappointed technological expectations -- Dude, Where's My Flying Car? -- but in retrospect, it's easy to see how hard it would be to meet the demands we actually put on them. In addition to, you know, actually flying, the Flying Car would have had to at least approximate many of the performance factors we get from today's automobiles: safety; ease of operation; carrying capacity; quietness; passenger comfort; and adaptability to existing infrastructure, most obviously parking spaces. Today, private aircraft meet none of these requirements. Even for many multiples of the price of an automobile, an aircraft is cramped, noisy, and very limited in the weight it can carry. (Full disclosure: I have a private pilot's license and enough experience in the Cessna 172 to know.) There are no collision safety standards for aircraft of which I am aware, since the weight added by meeting such standards would keep the aircraft on the ground. No suburban neighborhood would tolerate the noise of even a Cessna taking off from the street that ran in front of its houses, nevermind the sound of a helicopter, which are physically painful to listen to from even 50 yards away. And that's just aircraft. Add in the requirement that the aircraft also drive around like a car, Back-to-the-Future DeLorean-style, and the goal is even further out of our technological reach.

And so it will be with the sexbot. As a matter of first principles, the human race didn't come to dominate the planet by being easily tricked into frittering away its reproductive capacity on inanimate objects. So it follows that to appeal to the broad mass of humanity beyond a handful of fetishists, the sexbot has a steep hill to climb: it would have to be physically and functionally indistinguishable from a real person.

At risk of appear picky, I'm not remotely impressed by what I've seen so far. The Daily Star article has a lot of pictures (some NSFW-ish), and in almost none of them would I ever mistake the sexbot for a real person even in still photographs. Consider the first picture of the article's slide show: six females, standing and sitting in two rows; the caption stipulates that it's a mix of humans and dolls. Four of the females are obviously dolls. I'm not sure about the standing girl in the middle -- she's partially obscured -- but the seated girl on the right is, just as obviously, an actually human, vastly more lifelike than the dolls. That contest wasn't even close, and again, I could distiguish real from fake in a still photograph. The gulf between real an fake widens by an order of magnitude when the dolls have to move under their own power. And (I anticipate) the gulf widens expontially yet again when we touch them. So, no, I predict that these devices will not actually move like a real person, and actually feeling like a real person (since that's what would be involved) is even further beyond the technology.

So, why is the RealDoll (the subject of the article and the Cadillac of blow-ups) and other like firms getting all this press? Certainly, titillation is involved, but I suspect that the main reason is to lure in venture capital. My two cents: move along, Wall Street. This isn't the wave-of-the-future you're looking for.

Generation Y

Content Warning: Mathematics, Probability, Monte-Carlo Simulations

I was thinking of the academic paper Steve wrote about the other day, the one that documented the discovery that 60% of the current European male population carried the Y-chromosomes of only three men living 5000 (or so) years ago. This was represented as showing extraordinary reproductive capacity among those three men and their male descendants, but I was curious as to how much of that capacity could be explained by random variation.

For instance, imagine 1000 men have two children each. Assuming an equal probability of having a boy or a girl, the rules of probability require the expectation that 250 of those 1000 men have two girls; thus, their Y chromosomes are eliminated in the first generation. And 250 families will have two boys; thus, their Y chromosomes will be held by half the first generation. It logically follows that, given a sufficient number of generations, random chance will eventually leave only one of the original 1000 Y chromosomes.

The math gets highly iterative after this, and I wanted to see how the distribution of Y chromosomes would change over time. But I wanted to alter the model to account for a growing population. I also wanted to let the number of children per family vary randomly as well, and for this I chose a Rayleigh distribution.

So, naturally, I wrote a Monte-Carlo simulation in Matlab:

% Ygen:  Calculate remaining Y-chromosomes over generations
clear; close all;
N=10000;    % Initial Male Population
Y=1:N;      % Chromosomes numbered 1 to N
nGen = 40;  % Number of generations
mu = 2:.05:2.2;             % TFRs to try
topone = zeros(5,nGen);     % Share of Y held by top 1%
cdf = zeros(5,round(N/10)); % Cumulative Density Function of Y
tPop = zeros(5,nGen);       % Total Population
tY = zeros(5,nGen);         % Remainin Y
for i=1:5,   % Total fertility rate (TFR)
    sigma = mu(i)/sqrt(pi/2); % scale parameter for Rayleigh distribution
    M0 = Y;     % INitialize array of Y
    tM1 = N;    % Initialize total number of men
    for j = 1:nGen,
        tM0 = tM1;
        cC1 = round(random('rayl', sigma, 1, tM0)); % number of children
        cM1 = binornd(cC1, 0.5); % number of male children
        caM1 = cumsum(cM1);
        tM1 = caM1(end); % Total number of men in next generation
        M1 = zeros(1,tM1);
        % Calculate Y for next generation
        % M1 = repelem(M0, cM1); % Not available until R2015
        % Alternative for R2014 and earlier:
        M1(1:caM1(1)) = repmat(M0(1),1,cM1(1));
        p = 1;
        for k = 2:tM0,
            q = p + cM1(k) - 1; % Current stopping index
            M1(p:q) = repmat(M0(k),1,cM1(k));
            % M1((caM1(j-1)+1):caM1(j)) = repmat(M0(j),1,cM1(j));
            p = q + 1; % New starting index
        % End Alternative
        M0 = M1;
        tPop(i,j) = tM1;
        tY(i,j) = numel(unique(M0));
        disp(['TFR = ', num2str(mu(i)),', Gen ', num2str(j), ': ', ...
            num2str(numel(unique(M0))), ' Y lines among ', ...
            num2str(tM1), ' men.']);
        n = hist(M0, Y)/numel(M0);
        ns = sort(n,'descend');
        topone(i,j) = sum(ns(1:round(N/100)));
    cdf(i,:) = cumsum(ns(1:round(N/10)));
h1 = figure('Name','Top 1% Y Share');
title('Share of Y-chromosomes carried by top 1%');
ylabel('Y-share'); xlabel('Generation #');
legend('TFR = 2','2.05','2.1','2.15','2.2');
h2 = figure('Name','CDF of Y');
n = hist(M0, Y)/numel(M0);
ns = sort(n,'descend');
title('Cumulative Density Function of Y after 40 Generations');
ylabel('Y-share'); xlabel('Y');
legend('TFR = 2','2.05','2.1','2.15','2.2');
h3 = figure;
title('Total Male Population');ylabel('Population'),xlabel('Generation #');
legend('TFR = 2','2.05','2.1','2.15','2.2');
title('Remaining Y');
xlabel('Generation #');
legend('TFR = 2','2.05','2.1','2.15','2.2');

As you can see, I examined five values of total fertility ranging from a mean value of two children per family to 2.2 children per family, and measured several quantities over 40 generations from a starting population of 10000 men. Note that because the distribution of family size is (roughly) Gaussian, some families will have zero or one child, and some families will have more than two. This may not sound like much of a range, but over 40 generations, it makes a big difference in the total male population:

We can also see the elimination of Y-chromosomes over time:

We can see that the higher TFRs retain more of their initial number of Y chromosomes longer. This is likely a combination of the populations getting bigger faster, making it progressively more difficult for random chance to produce an all-girl generation for a single chromosome, and also a relatively lower probability that any given family would produce only girls, given that they get more chances for a boy.

Now let's look at the cumulative density functions for the Y-chromosome distributions:

For instance, we see that with a TFR of 2 (and therefore a constant male population around 10000), only 500 or so of the original 10000 Y-chromosomes remain after 40 generations. In contrast, for a TFR of 2.2 (and a population approaching a half-million), the top 500 Y-chromosomes are only carried by about half the male population; the other half carry one of the other 1700 Y-chromosomes still in existence.

Finally, let's look at how the share of the male population carrying the top 100 (or 1% of the original) Y-chromosomes changes over time:

In contrast to my other graphs, this data is noisy. Better graphs could be generated with more lengthy simulations. Feel free to borrow this code and try it yourself, and if you have the 2015 version of Matlab or later, you can avail yourself of the repelem() function, which makes these simulations much faster. In any case, here we see that as the population gets higher, it becomes increasingly difficult to eliminate Y chromosomes from the gene pool. We also see that in none of these models do even the top 100--let alone the top 3--chromosomes account for 60% of the male population after 40 generations.

Of course, 40 generations is only, what, 1000 years? Not 5000 as per the paper. And I will speculate that the reproductive history of Europe has been characterized by much higher TFR rates than those I have examined, punctuated by war, pestilence and plague, and thus plenty of opportunities for Darwinian struggle. But certainly, multi-generational differentials in fertility are necessary to account for the observed Y distributions among the present numbers of Europeans.

Let's rerun the above code for N = 1000 and nGen = 100 to produce this:

So, if I keep my male population low enough (1000) for long enough (100 generations), I can (at least on this particular simulation) get the top three Y chromosomes up to a 46% share of the population. But look at this graph from the same simulation:

Note that with a growing population (TFR = 2.2), between generation 50 (population 131K) and generation 100 (population 15.6M), the number of extant Y chromosomes stabilized at 216 from the original 1000. Not one single Y chromosome was eliminated in over a thousand years of procreation. So while it is theoretically possible that random chance would eventually eliminate all but one of the original 1000 Ys, my expectation is . . . well, the sun will probably have fizzled out by then.

The Great Fat Liberal Civil War

Via David Codrea, I read a Facebook commenter:

Oh haha. I love this shit. Haven't even read the linked article. It's all the BS 'gun owner' bravado and jingoistic 'blades of grass' enthusiasm that gets me. Most 'gun owners' wouldn't do shit. Most are fat untrained fudds. You talk about 1776 but those guys were not burger munching blimps! This has nothing to do with whether or who anyone would win. It has everything to do with Facebook Meme Patriotardism, backed up by nothing by untrained bloviating fat fucks. [Emphasis added.]

This reminded me of an episode from the NPR show This American Life from last June, "Tell Me I'm Fat" (transcript here), the upshot of which was to agitate for "Fat Acceptance". It featured several fat or formerly fat people telling their tales of suffering societal discrimination. (I'm making this sound more lame than it actually was; TAL pushes a hard-Left agenda, but it sounds reasonable in the NPR style.)

From Act IV of the transcript:

It's so common to judge people on their weight. And of course, so often there is this moral dimension to it that is just gross-- this idea that you're fat because you're weak, you can't get control of your own life.

Today on our program, we're saying maybe that is not the most accurate or the most helpful way to look at this. This next story is about a very specialized example of this kind of moralism. You may know that there's a Christian weight loss movement. And it's big, with seminars and books like Help Lord, the Devil Wants Me Fat.

Let me interrupt for a moment. First of all, don't read too much into this. Almost everything in popular culture has a "Christian" version. We have books on Christian dating. Christian financial management. Christian rock music. (Come to think of it, the only thing I haven't seen is Christian porn, but I'm still looking.) So it shouldn't surprise us that we see Christian weight-loss. These books and seminars are written and read, presented and consumed, by Christian and secular audiences for the same reason: people want to not be fat. Their existence do not represent some kind of deeply-rooted theological position on obesity.

But TAL finds an anecdote anyway, quoted here at length:

[Physical Education instructor Paul Brinson] got a call from the brand-new Oral Roberts University asking him to head up their new phys ed program. Paul was a Pentecostal Evangelical Christian. And Oral Roberts was maybe the most famous televangelist and faith healer in the world. Two of the biggest things in Paul's life were coming together-- God and exercise. So Paul and his wife moved out to Tulsa.


Oral Roberts wanted to teach a lifestyle to his students, what he called the whole man philosophy-- mind, body, spirit. They would all have equal importance at the school. Here's how the provost from back then, Carl Hamilton, explains it now.

Carl Hamilton: Our bodies are the home of the Holy Spirit. Making that home a fit one is one of the ways to glorify God and the Holy Spirit.

Paul would be in charge of making that home a fit one. In the fall of 1974, he took over the school's aerobics program at a brand new $2 million center with a fancy indoor track. Every student had to get a certain number of aerobics points per week. They might get a point for walking a mile or two for playing doubles tennis or five for bouncing on a trampoline.

Remember, this is 40 years before Fitbits. Back in 1974, just the idea of regular exercise was cutting edge. Jogging had only just become a thing. Aerobics was brand new. And here was ORU with all its students running around in headbands and tube socks. Then Oral Roberts had another idea-- to push the fitness program even further. Here's Paul.

Paul Brinson: At the graduation ceremonies of 1975, at the end of my first year there, Oral Roberts observed several students who graduated that he saw were obese. And within the next few days, he contacted Dr. Carl Hamilton, the provost of the university-- and this is, I understand, how this took place-- and said, I really don't want to see significantly overweight, obese students graduating, because it indicates that they have not met the goal of the university to be physically disciplined. And please develop some guidelines or some criteria so those students make progress or do not graduate. Maybe do not graduate is too strong a term, but that, anyway, something is done about that.

In that moment, something changed. Paul's fitness program wouldn't just keep tabs on students' exercise. Now he'd make sure they weren't fat. Suddenly how you looked mattered.

Personally, none of this sounds particularly scandalous. On the contrary, I spent both college and the first 20 years of my professional life living under physical fitness standards. The specific requirements have come and gone over the years, but Body-Mass Index (BMI) measurements have always been among them. And the last few years of active duty included actual waist measurements.

The military's interest in maintaining physical performance requirements among its soldiers and airmen is pretty obvious. And to the extent that actual data show correlations between being overweight and healthcare costs, then it follows that keeping troops at a healthy weight is worthwhile financially. But these are not the service's only reasons. They have never made a secret that "professional appearance" is one of the goals as well. For instance, the Army's Officer Evaluation Report (OER) has a single rating category for "fitness and military bearing".

So while it might surprise me to see a civilian institution take an interest in how its students' appearance reflects on their school, it doesn't strike me as especially heinous. But my objection is that while the "Pounds Off" program is an interesting bit of ORU folk history (although the school still has a physical fitness requirement -- now monitored by FitBits -- the "Pounds Off" program was discontinued in 1978), it has almost nothing to do with the battle over fatness in the current decade. Forty years ago, when America at large was just coming into awareness that the new era of cheap calorie abundance would require lifestyle adaptation, body type didn't have a class dimension. Today, it does.

Let's look at another example from the program, a conflict over fatness between radio host and author Dan Savage and blogger Lindy West:

West: Then Dan wrote a blog post entitled "Ban Fat Marriage." He was responding to some Republicans' argument that gay marriage should be illegal because gay people supposedly die younger.

Here's Dan. Quote, "Why stop with gay people? Iowa should ban fat marriage. There are, according to the state of Iowa, more than 1.4 million obese people living in Iowa. The social costs of Iowa's obesity epidemic are pretty staggering. And those costs include premature death and lower average life expectancies for Iowans."

In response, I threw up a quick blog post. "Hey, Dan. So now that you're equating the stigmatization of fat people with the stigmatization of gay people, does that mean you're going to stop stigmatizing fat people on this blog?" Nothing. I waited a few days. Nothing.

No exchange better encapsulates how the cause of fatness is strictly a civil war among liberals. On the one hand, Dan Savage wants to weaponize fatness in his status competition with the lower classes in Flyover Country. The appeal of this is obvious: the inverse correlation between body weight and social class among women is well documented from the General Social Survey, probably among other sources. (The record is more mixed in the case of men; lower class men show greater body variance -- more fat and more skinny -- while upper class male body types are more clustered towards "average".) Meanwhile, a large (ha!) faction of feminism -- and if you read the transcript, all the fat people weighing in (see what I did there!) in favor of fat acceptance are women -- want for fatness what homosexuals have: a seat at the progressive table as a designated victim group.

Since TAL, at least in this context, is taking the side of feminism, then as a rhetorical strategy, trying to link the public push for weight control to an ORU program forty years in the past looks like a smart move. But we should recognize that strategy for what it is: just rhetoric, and having no basis in present reality. Neither I nor Christianity have a dog in this fight, and none of the belligerents are on my side.