Twitter uses an algorithm to figure out your gender


#24

Have you ever deliberately clicked on a promoted tweet? How would they determine your shopping pattern if you haven’t? The only time I click an ad is in google search results when I hate that company and want them to give google money.


#25

This is misleading scaremongering. They disseminate your individual information? No they don’t, they allow advertisers to reach targeted demographics. That doesn’t mean the advertiser has any fucking idea who you are (unless you then click the ad, and then buy their product) and none of these companies present a library of users through which advertisers can peruse our profiles. None of the companies you listed, to my knowledge, on sell their user marketing profiles, because that database is what makes their money for them. They sell access to demographics, not information about you.


#26

Do marketers care about gender or sex?

What is gender?


#27

I am dissapointed in the responses to my critiques. I am wary of the modern replacement of democracy with a technologically fueled form of authoritarian capitalism. I was hoping to provoke thoughts on the ease at which the behavior of silicon valley corporations are de facto reshaping how we engage socially. I was hoping to discuss the larger cultural ramifications of the particulars discussed in the original post. It seems my opinions have sparked negative responses. I am the first to admit that I am a Luddite. This has been a learning experience that has taught me that my viewpoints are unwelcome amongst some members of the tech community here at the boingboing message boards.


#28

No, you just have to learn to be on topic. There are a thousand places you could talk about this particular issue, but nobody likes a hijacker. Find the appropriate place to discuss, not just one that’s convenient to you. That’s selfishishness.


#29

Just start a new topic. And that way it won’t suffer from the 5-day guillotine, either. Wins all round! :smiley:


#30

I’m puzzled: the article claims that there are roughly equal numbers of men and women using Twitter, but in every example cited, there are significantly more men than women. Is that just a fluke?

I don’t have that many followers, just 41, so it was relatively easy to look over them. Twitter said, 10 women, 30 men. By my count, 11 women, 19 men, 9 organizations, 2 unknown.


#31

Dude. You opened the thread by accusing a well respected contributor of writing an undisclosed paid article, then failed to justify that accusation quite spectacularly.

If you keep hanging around you’ll find plenty of opportunities to engage on the kind of subject matter that seems to interest you, but this isn’t it. This is about something else.


#32

Here’s one possible explanation:

Say there are 10 women on twitter and 10 men. However, each man follows 6 other people and each woman follows three. There’s still the same amount of men and women, but everybody has twice the number of male followers.


#33

Organizations and corporations are not only people these days but they are also male?

Brings up the interesting questions about how organizations interact with twitter vs how twitter analytics work.


#34

I have a strange feeling Twitter’s gender algorithm is broken.

I also have a feeling that it may be deciding bots are male. I know I have a lot of bots following me.


#35

Maybe if they have someone geolocated at a Final Sale and typing in all caps, it’s time to put that person in gender lockout for a bit? Lots of little bins. Maybe I can pop up soygendered sometime if I get on an HDR foodie bender. Looking forward of course to the ‘ads made my circles work publicly’ defense pulling boffo box over corpse drama serials. That’s the baking contest Father Brown Mystery and Eraserhead mashed up, right?


#36

Maybe it doesn’t really matter if they do or don’t, I kind of get the feeling that it only matters that they get other people to think they can target a demo, if at least, better than others.

In this sense, twitter can say, in business speak, that they know who they’re users are and therefore, how they should potentially shop.

After all, its not like most businesses really do rely on scientific data to operate, or that ad networks have an impeccable record in targeting “demo’s”, they all sort of realize that its the best they can do at the moment while at the same time publicly presenting themselves as 100% confident about their business models and results.


#37

I only have about 70 followers, so I did a manual count. Twitter estimated ~40 men and ~30 women. The actual figures are 16 men and 35 women; the rest are a mixture of nonbinary people, bots, plus two humans whose gender I’m not sure of. So essentially the algorithm assigned ALL the bots and the nonbinary humans as male, in my case. (As some of the bots are, heh, decidedly female-presenting, it seems likely that they assigned some human women as men).


#38

I only have 55 followers, and I know most of them personally, so I decided to check Twitter’s accuracy by doing a manual count. Twitter claims that 91% of my followers are male and 9% are female. My own count was:

male: 35 (65%)
female: 12 (22%)
nonbinary: 2 (4%)
bots and rss feeds: 5 (9%)

Most of my female followers clearly identify somewhere in their profile that they’re female, and/or have a female name, so I really don’t know why Twitter is undercounting them.


#39

This topic was automatically closed after 5 days. New replies are no longer allowed.