I’m a trans woman – here’s why algorithms scare me | Hacker News Reader

I’m a trans woman – here’s why algorithms scare me

dazeddigital.com

8 pointsamyjess13 comments

> These false ideas about gender have been thoroughly debunked by science.

As much as I wish that were true, it's not. And not only does claiming it to be solid science not change anything, but also it weakens the standing of the person saying it. And by extension, the rest of us who live as we feel we need to live, but not at the expense of truth and intellectual honesty.

Can you clarify which ideas are scientifically disputed?

I would guess they refer to the suggestion that gender is determined by the genitals that a person is born with, which the parent comment think is correct. Just like the article however, I agree that science has debunked that theory. Androgen insensitivity syndrome for example prevent the masculinization of male genitalia in the developing fetus with male phenotype. They are described by science as XY female, but does not have ovaries or a uterus.

There is also studies done on neuroscience of sex differences, which found that transsexual individuals tend to have the brain characteristics and structures that do not match the genitals that the person was born with. Again a rather strong indication that gender and genitalia do not always match. There is quite a lot of research that indicate transgender as caused by biology, where individuals get a mix of both male and female traits, and genitals in those cases is simply one of many traits.

I gave you an upvote because this is all important information and likely helpful for other readers, but unless I've misunderstood the parent commenter I'm not sure this is what they meant. All of this information is consistent with the article author's claims, which the parent commenter disputed.

Yes. I was a bit unclear but I was trying to addressing that the top parent comment disagree with the article and think that genitalia determine gender and that there is no science that disproves it. The article is right however that science has pretty much proven that gender is a multitude of traits, and as with any other collections of biological traits those does not always align in every individual.

I edited the post slightly to hopefully make it more clear.

Well first of all, she misquotes the memo, which the New York Times reported as coming from the Department of Health and Human Services — they say 'sex', not 'gender':

> “Sex means a person’s status as male or female based on immutable biological traits identifiable by or before birth,” the department proposed in the memo, which was drafted and has been circulating since last spring. “The sex listed on a person’s birth certificate, as originally issued, shall constitute definitive proof of a person’s sex unless rebutted by reliable genetic evidence.”

I'm not saying I agree with the memo. But there's a reason to use the word 'sex' instead of 'gender' — the former refers to ones biological sex, which in humans is essentially determined by whether the person has a Y chromosome.

There is NO strong body of evidence indicating that people who are trans are that way because of biological factors. There are a few small studies, but nothing that, quantitatively speaking, justifies her claim that "[...] false ideas about gender have been thoroughly debunked by science".

There is actually a great amount of contention in academia about gender. You really needn't look further than the debate over whether gender is a social construct — a notion popular with some feminists and in the social sciences. The fact that it's so strongly asserted by as many people in the social sciences as it is, despite overwhelming evidence to the contrary.

A specific example: there's the "Social Role Hypothesis" which basically postulates a evening out of the distribution of men and women as a society becomes more industrialized and egalitarian [0]

However, a study [1] conducted by Armin Falk, an economics professor at the University of Bonn, and Johannes Hermle, a doctoral student in economics at the University of California, Berkeley of 80,000 people in 76 countries strongly disfavors that hypothesis.

The study supports quite the opposite — as life becomes less hard in a society, and as it makes progress in terms of gender equality, men and women tend to manifest MORE gender preferences.

What we have at the moment are deep divisions in academia, where certain topics are not being approached in an rational, intellectually honest, evidence-based way. Frankly, it approaches dogmatism.

When so much evidence suggests that biological sex and gender are very much _tightly_ interwoven — they are absolutely causally linked, and you still have people in academia claiming that they're not — even to the point of saying that gender has no biological basis — that's not "thoroughly debunked" by any stretch of the imagination.

I don't blame the author. She probably has sources that one would take as authoritative at first glance. The problem is that it's not backed by a thorough, honest, cross-disciplinary discourse. And sadly, for people like her and I, it won't be resolved until everyone puts their feelings aside and resolve to actually get to the bottom of it, regardless of whether the truth resembles what we wish were so.

[0] https://www.researchgate.net/profile/Mary_Kite/publication/2... [1] http://science.sciencemag.org/content/362/6412/eaas9899

> At the University of North Carolina at Wilmington, researchers created a system that attempts to identify transgender people before and after they medically transition with hormone replacement therapy. To do this, researchers scoured YouTube for ‘transition timeline’ videos, which typically involve a series of photographs charting a person’s face changing over time. These videos are extremely personal, and a source of empowerment for the trans community...

This isn't a trans issue, this is a 'digital consent' issue. As noted in the linked Verge article, "[t]he MegaFace dataset compiled by the University of Washington, for example, contains 4.7 million images of roughly 627,000 individuals — all taken from Flickr users." It is fair to say that many of those photos are deeply personal.

So then, this raises a much larger question - not about digital ethics as applied to trans people, but digital ethics as applied to people: where are the lines drawn in fair-use for public content? Do you get to dictate what happens to your data that you upload to someone else's server publicly?

I find it entirely reasonable to hold mass aggregation of content to different standards than use that's either individual or low-context.

An algorithm that trys and detect a male or female is not bad or unethical period. Gender != sex, males and females have different bodies and faces based on different genetic makeup between XX chromosomes and XY chromosomes.

> If the researchers had consulted any trans people prior to beginning the project, they would know that many of us transition because we don’t want to be linked to our past name or appearance. We want to live our truth in the present and define our own future – not be algorithmically chained to false identities we were forced to wear in the past.

The concern doesn't seem so much to about the existence of the algorithm, but instead it's uses.

I think that the best solution to this problem is to ban gender-based discrimination by advertisers and data-aggregators.

These algorithms only exist because marketers want to buy impressions from "25-34 year old men" If that was not a legal option, these algorithms would not have an economic incentive to exist.

Surely "people who like power tools" as a demographic is much more targeted, both excluding "men who don't care about power tools" false positives and including "non-men who do" false negatives. A system of advertising that allows gender based targeting inherently reinforces gender stereotypes.

Targeted ad placement is much more valuable for advertisers and those who profit from ad placement. Surely there are use-cases that warrant a sex-based discrimination: sports bras, as an example, or hair pomade, or gendered magazines (i.e. Men's Fitness; Seventeen).

You're certainly right that advertisers and advertising platforms have little to no financial incentive to mitigate the externalities of reinforcing stereotypes by segmenting demographics on the basis of protected class distinctions.

However, I think as a society, we should discuss whether these externalities are worth the value that advertising provides to society.

Reply on news.ycombinator.com