Wokeness and ethnicity in artificial intelligence

It is unclear whether Google’s AGI moonshot subsidiary Deepmind is as wokefied as Google itself. Judging by recent blogposts „Strengthening the AI community“ and „Causal Bayesian Networks“ they are not far off. „Strengthening the AI community“ is about including members from underrepresented groups into AI research, despite the reason for the underrepresentation being that very few member of these groups show the ability to contribute to top research.

„Causal Bayesian Networks – a flexible tool to enable fairer machine learning“ is potentially much more sinister than that. It basically describes a tool to introduce ideological biases into machine learning models. To quote from a figure caption: „Figure 2b: In the second scenario, female applicants apply to departments with low acceptance rates due to systemic historical or cultural pressures, and therefore the path G —>D is considered unfair (as a consequence, the path D —>A becomes partially unfair).

Of course whether this is the case „requires expert knowledge“.

Open AI – the other big outfit stating AGI as the explicit goal and actually producing ground breaking research plays a very similar tune. Its Open AI scholars – a mentored internship of sorts – have the key qualification of not being white men.

These tries to increase diversity in AI come with the expressed goal of bringing many more people to the table in a future where AI or even AGI has a huge impact on how the world is run. In the extreme case a superintelligent machine would be created that is imbued with certain values and due to its vastly superior intellect starts calling the shots.

The values chosen to be imparted on this machine, if such a thing is even safely possible, are supposed to be representative of all of mankind. And not just a small subset of white men, who for some reason seem to always be the ones to create steam engines, cars, nitrogen fixation, airplanes, nuclear bombs, antibiotics, computers and maybe finally also AGI.

When this topic is brought to the table I always wonder whether this is just completely politically motivated or whether they actually believe this to be a sensible idea. I mean, let’s assume you have certain values, for example female emancipation, liberal democracy and the scientific worldview. And you are about to create an AGI that will make sure that the arc of history will bend towards the values it is seeded with.

Now, if you bring other people with other values to the table you are going to have to make compromises. Female emancipation, yes, but not for Saudis. Liberal democracy, ok, but uploaded Putin will be Russian Zar forever. A scientific worldview only inasmuch as it doesn’t conflict with various religious dogmas.

I only see two reasons why somebody would honestly propose to bring in lots of other people to figure out the values by which the future will be built.

Either they think that their values are self-evidently correct and everybody else will fall in line. In which case, A) they are wrong and B) why bring them to the table at all? That’s just an empty gesture.

Or they are cultural relativists and honestly believe that other peoples values are just as valid and good as their values. Which of course means that they don’t have any values.

Often it will be a mixture of the two enabled by muddled thinking. This is especially clear when the people with different, but just as good or even better values, are future generations. Here, the argument is being made that locking in certain values by seeding a superintelligent machines with them is a horrible thing, because it doesn’t allow future generations to develop their own set of values.

Proponents of this argument seem to imagine that future generations will be wiser and nicer than we are and that all value differences are consequently going to be of the variety that if confronted with a well-argued version of the new values we immediately understand that we are wrong (we kind of deep down knew it all along).

In this case I would argue there wasn’t really a value difference. Just possibly deeper understanding or clearer thinking.

They never seem to take the possibility into account that future generations develop into a less benign direction. Maybe they begin to see the benefits of slavery, cannibalism and genocidal warfare. If you think we should program the AGI to avoid these directions, than you don’t really believe that it is horrible to lock the future humanity into one set of values.

To bring this full circle let’s take a look at OpenAI’s company outing. Employees together with significant others, possibly family members as well. You will find some Indians, quite a few North-East-Asians and a lot of white men.

Open AI

Dominance is build into the bell curve

In sports we often have incredibly dominant athletes, that for a while make it very clear who is on top. In chess that is particularly noticeable with Carlsen topping the rating list for almost ten years now, Kasparov before him being head and shoulders above everyone else and many other champions enjoying long undisputed reigns (Karpov, Casablanca) or at least short extremely dominant stretches (Alekhine, Fischer).

But the same holds for many other sports as well. Federer, Klitschko, Tiger Woods, Carl Lewis, etc. etc. I recently realized that this is at least partly a direct consequence of abilities being normally distributed. In a normal distribution the number of athletes within an ability bucket of a certain size drops exponentially the farther out from the mean you look. Intuitively this means that the top athlete might get his very own bucket, while the preceding bucket is already filled with, let’s say, ten rivals. Consequently the average distance to the next best athlete must be exponentially smaller the closer you get to the mean.

This exponentially bigger distance to the next best rival is what we call dominance. Counterintuitively this entails that the stronger the competition in a given sport the more dominant the top athlete is likely to be. Simply because the top in a very competitive field is going to be farther out from the mean. Of course this mathematical relationship doesn’t hold as strongly in teams sports. In teams sports dominance is more likely to result from winner-takes-all dynamics.

Random Thoughts – The Problem with Polyamory

I
Modern dating apps and sites have given us data to quantify differences in mate choice between men and women. Specifically, female hypergamy, that is the preference of women to date the highest status men, can now be observed in hard data. It turns out that average man is really not attractive to the average women. Instead 80% of the women chase after 20% of the men. This is in itself worrisome if one regards falling birth rates and disenfranchised men as a problem.

II
It is also the reason why I think polyamory would be a very bad social norm. Polyamory is the habit or norm of having multiple partners at the same time. Its proponents argue that it would solve the disenfranchised men’s problem, because they could still become secondary or tertiary partners. Why women would want a secondary or tertiary partner that is not attractive to them is anybodies guess. Instead polyamory solves the problem for the female conundrum: Suddenly the attractive 20% of the men can be available to 80% of the women.

III
In the end polyamory makes the mating market more efficient. Just like dating apps or dating sites. Unfortunately, making the mating market more efficient just means that the messed up female preferences make life miserable for both men and women.

Random Thoughts – White Supremacy

I
Bloggers like HBDchick and increasingly mainstream scientist and authors argue that the unusual history of (partly Church mandated) outbreeding, certain societal structures / customs and the steady replacement of the lower classes by the middle class has changed Europeans genetically to exhibit higher intelligence, more moral behavior induced by a stronger propensity for guilt, less nepotism, less violence and more empathy, altruism and trust for strangers than basically any non-European ethnicity. This is the putative origin of the WEIRD psychology.

II
I don’t know to which degree this is correct, but it is definitely not a non-starter. One can leverage a lot of data to show the trait differences of WEIRDoes compared to the rest of the world. There is some data to imply a genetic cause of the observed differences and the evolutionary stories at least sound somewhat plausible.

III
Now HBDchick is likely a lovely person with no bad intent, but her work can of course be used to argue for some form of white supremacy. This, however, quickly leads to a funny contradiction, which seems to me to be the central contradiction of white supremacy:
Along every dimension white supremacists are more like the rest of the world and less like the WEIRD.

IV
They are less intelligent, to a degree that the right generally has a recognizable human capital problem. They abhor the „pathological“ altruism exhibited by their co-ethnics. They are certainly more violent and would have a hard time to argue for moral superiority. One could go so far as to say that Western leftism is exactly the quintessence of what makes Europeans different from the rest of the world!

Seven bad trends

In a Q&A a few years back, Jordan Peterson opined that there are 6 or 7 bad trends, which if unchecked could lead to very bad things. I’m paraphrasing, I don’t remember the exact words. But I would have loved to get the opportunity to ask him which trends he was talking about. The picture of seven bad trends stayed with me and occasionally I try to come up with my own list.

Following the scientific and technological progress in fields like AI, genomics and spacefaring closely, can leave the strong impression that mankind is quite close to making very big strides. Strides that would take us to a position from which our current problems are (easily) solvable.

I call that „reaching escape velocity“. The point where our technological and scientific acumen accumulates faster than our problems compound.

Of course when you are sitting in a rattling box of metal that accelerates into the impenetrable mist ahead, it is very unclear whether you are ever going to reach escape velocity. Maybe you’ll crash and burn instead. Or just sort of putter out.

Climate change can serve as a stand-in for all kinds of environmental degradation and resource exhaustion. Mass migration and birth rate collapse together lead to a population turn-over in the countries that historically have been the engine of scientific and technological growth.

This alone might stall the global growth machine, but on top of it there is dysgenic decline going on in almost all countries at the rate of probably 1 IQ point per generation possibly more. That the ongoing population replacement also engenders ethnic conflict only makes it more likely that the Western world will not be able to continue to drive innovation.

It is unclear how much automation already plays a role. We can also use it as a stand-in for all economic forces that squeeze the little man. So far mostly globalisation. Next to the population replacement this is the other big driver of political polarization. Which ultimately feeds into ethnic conflict. On the one hand because almost all violent group conflict is between ethnic groups and on the other hand because very concretely left wokeness is largely based on ethnic hate and the right-wing reaction to it is no stranger to racism either to put it mildly.

Random Thoughts – Discrimination

I
The investigation into a lot of societal questions comes down to base rates. If a black guy is called a fucking n***r, in many cases this undoubtedly comes down to racism. If he and other black persons are often insulted in such fashion one can probably make a reasonable case for a racist society.

On the other hand I was called a fucking bastard just last week for no reason. Would that person have availed himself to a racist insult if I’d been a darker shade of pale? Quite possibly. Insults are meant to hurt and racism fits that bill.

So a better measure for the prevalence of racism in a society would probably be whether minorities are more often insulted than people of the majority all else equal and not so much how those insults are couched. People who hate everybody equally are not racist.

II
I wonder whether displays of homophobia, especially among adolescents, aren’t often better explained by status games than, well, actual homophobia. If gays are lower status, implying somebody is gay, or denying to be gay is not motivated by homophobia. It is merely a try to lower somebodies status or to defend against such an attack.

Of course, one could say that gays being lower status is due to homophobia. But that is not a foregone conclusion. There are systematic trait differences between gays and heteros, so gays might be lower status because they are more feminine. Or they are lower status because they miss out on one key aspect of intra-male competition. In that case a teenager calling somebody gay, is just saying „you don’t compete for females!“. And the forceful denial is just „but I do, I do“ bangs chest.

III
The point of these speculations is not to explain away suffering caused by hurtful remarks or actual discrimination. But in a way I do discount lived experience. Namely, when it is invoked in lieu of actual solid evidence for societal problems.

The most ethnic names

Because counting names is one of my modus operandi, today we are going to take a quick look at the names that are really specific to each ethnic group in the US according to the US census. To that end we order the names by percentage of the population that belongs to the given ethnic group and bears that name. We restrict ourselves to names borne by at least 1000 people.

For whites these are

99.65 NEWSWANGER
99.46 BORNTREGER
99.4 WAGLER
99.15 BORNTRAGER
99.14 MEISELS
99.1 BEECHY
99.09 BORKHOLDER
99.03 SHEVCHENKO
99.0 STOLTZFUS
98.99 SWARTZENTRUBER
98.87 NOLT
98.86 HERSCHBERGER
98.81 ESH
98.71 KUSHNIR
98.63 HODZIC
98.59 IVANOVA
98.58 STASKO
98.54 DWORAK
98.54 BALLWEG
98.52 ROHRBAUGH
98.52 PUSKAR
98.52 KOVALENKO
98.44 LENGACHER
98.39 BARHORST
98.38 HUETHER
98.37 REINEKE
98.34 PETERSHEIM
98.3 WENGERD
98.29 FLATEN
98.25 GEERS
98.23 HOCHSTETLER
98.21 SARGSYAN
98.21 HISSONG
98.2 GUREVICH
98.2 BERKEBILE
98.19 SCHMUCKER
98.19 BYLER
98.13 KATS
98.12 SOBOLEWSKI
98.11 KACZYNSKI
98.11 CHOINIERE
98.09 GOODLING
98.08 HERSHBERGER
98.08 BEACHY
98.04 IVANOV

Germans, Dutch, Eastern Europeans of different couleur including Ashkenazim. Latecomers that didn’t have the opportunity to gift their names to the African American population.

For African Americans the top names are:

97.75 KOROMA
97.7 ACHEAMPONG
97.46 AGYEMANG
97.37 WARSAME
97.21 MEKONNEN
97.06 ANYANWU
96.78 ALTIDOR
96.75 BOATENG
96.69 FRIMPONG
96.63 OSEI
96.62 BALOGUN
96.6 OKAFOR
96.59 NJOKU
96.56 NWOSU
96.5 TESFAYE
96.5 OPOKU
96.49 YEBOAH
96.46 JALLOH
96.4 ASSEFA
96.2 ADJEI

Same pattern applies. These seem to be genuinely African names. Maybe some latecomers also. If we increase the cutoff to 10.000 people bearing the name, we get 90.49 SMALLS, 87.53 WASHINGTON, 86.74 PIERRE, 82.86 MUHAMMAD, 80.85 HAIRSTON, 80.4 RUFFIN. All the African names are gone and the names become very unspecific.

98.82 VOONG
98.63 ZHEN
98.38 KUANG
98.26 CHUONG
98.25 XU
98.23 ZHU
98.21 QIU
98.2 ZHOU
98.19 XIE
98.11 ZHAO
98.07 XIONG
98.06 ZHANG
98.05 QIAN
98.02 ZHUANG

Boring. Yes, East Asian Names are common and really specific. It takes a while until an 97.54 BALASUBRAMANIAN comes along.

98.84 LOPEZPEREZ
98.8 GARCIARAMIREZ
98.75 CARACHURE
98.72 GARCIAMARTINEZ
98.64 EQUIHUA
98.63 VILLAFAN
98.61 HERNANDEZLOPEZ
98.53 GONZALEZLOPEZ
98.5 PEREZHERNANDEZ
98.44 RAMIREZLOPEZ
98.39 PEREZPEREZ
98.38 SANCHEZGARCIA
98.36 GARCIAPEREZ
98.28 MARTINEZLOPEZ
98.26 LUCATERO
98.25 GARCIAHERNANDE
98.23 OREGEL
98.22 SOLACHE
98.22 ALAVEZ
98.19 CUAUTLE

The hispanic names are dominated by the Spanish style double name. Makes sense that these are recent immigrants still following that naming convention. If we increase the cutoff to at least 10.000 people bearing the name, these double names completely vanish.

95.65 BEGAYE
95.0 TSOSIE
94.56 YAZZIE
94.3 BENALLY
93.84 BEGAY
92.13 NEZ
76.52 HARJO
72.9 LOCKLEAR
72.55 BAHE
71.85 POITRA
68.37 JIM
67.77 AZURE
63.82 SHORTY
63.68 OXENDINE
57.85 CHARLIE
57.51 BELGARDE
56.18 CHARLEY
52.6 TSO
51.4 CHEE

And finally we have a really short list of native American names. There just aren’t all that many Natives around and most of those are mixed in some way or form, which means that their names have likely bled into the white population.

These investigation shows the difficulty of assessing ethnicity by name: Only recent immigrants still bear really distinguishing names.