Slashdot Mirror
Tech Firms Have An Obsession With 'Female' Digital Servants (zdnet.com)
An anonymous reader writes: Alexa, Tay, Siri, Cortana, Xiaoice, and Google Now. These technologies all have one thing in common -- they are digital servants aimed at a mass-market audience that feature a "female" voice or persona. And it's not just the voice or persona of the digital persona we interact with that is biased. The results of those interactions also demonstrate male favoritism. It took Apple more than four years to fix Siri's responses to questions about abortion services, and yet the company didn't seem to have any problem programming Siri to search for prostitutes and Viagra. Here's the gender breakdown for the tech workforce of each company:
Microsoft: 83.0% male, 16.9% female
Google: 82.0% male, 18.0% female
Apple: 79.0% male, 22.0% female
Amazon: 61.0% male, 39.0% female
Microsoft: 83.0% male, 16.9% female
Google: 82.0% male, 18.0% female
Apple: 79.0% male, 22.0% female
Amazon: 61.0% male, 39.0% female
Female voices are easier to understand. This has been known for decades. It has nothing to do with exploitation. As if you could even exploit a computer program in the first place.
Siri also has a male voice option. I have a friend and that's the option she chose.
Another hypothesis: this is an imaginary problem, since Siri debuted with a male voice in other regions (e.g. UK and France) and is capable of being changed to either male or female voices in all or almost all major regions. I haven't checked, but I'd assume the same is true for Google Now and that Microsoft and Amazon are likely working on the same thing too. At best, there may be an argument that this is an American cultural issue, but suggesting it's a general tech industry problem would require that we ignore the obvious evidence to the contrary, namely, that these products aren't female by default for all users.
Nice of the summary to skip over that and go straight to SEXISM
Amazon's actual numbers here: https://images-na.ssl-images-a...
"PROFESSIONALS" = 74.5% male
"TECHNICIANS" = 88.8% male
"LABORERS & HELPERS" = 54.6% male
Actually most automated voices are female because a female voice is easier to hear against background noise. Outside of the traditional tech area from this article I can think of two more I hear on a daily basis. The parking ticket machines where I park my car and the next station announcement on the public train that I catch. Male, deeper, voices would be much harder to hear against the background rumble of the city. This applies to any technology where a use case is mobile.
Its the use-case that matters. All these assistants are 'speak' close to your ears. Your ears also receive predominantly 'bassy' background noise, as you mentioned. A higher pitched sound is different enough to be more easily isolated from the background buzz.
Thats exactly why hunter-gatherer languages are full of high pitched tongue clicks: prey cant hear the hunters talk because the sound stops at the first tree.
So high pitch is the frequency range we have evolved to communicate with.
- both the ability to express (clicks, consonnants and hisses)
- and the ability to hear (our ear do cover the necessary range)
Explain me again how this is an argument, against using female (higher-pitched) voices ?
How your explanation of physics contradicts the parent poster that higher-pitched voices are better heard ?
Ever walk past a nightclub ? Notice how you can usually hear the bass drum through the walls but not the rest of the music ?
(BTW, your explanation is incomplete. The dominance of basses isn't only due to the diffusion being proportional to 1/f . It's also due to the way how different material conducts different frequency ranges. Few night club are completely in the open).
So you've (more or less) successfully demonstrated that most noise that get the farthest and dominates the most is low-frequency.
Given all this low-frequency noise, please explain me again why you think that the parent is wrong in proposing that a high pitch voice (say a soprano - high range female) would be better heard against such low-freq noise than a low pitch voice (bass - low range male) ?
If you are going to attribute something to the laws of physics you should probably know what they are first.
If you're going throw around science in your reply, maybe you should pay attention of this peculiar field called Psychoacoustics.
It's science, it works and it's what nearly every modern audio compression algorithm runs on.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
Unfortunately you've been bitten by the progressive non-scientific but called Gender Neutrality which views both genders as being the same in all areas, and blaming of culture for corrupting them into gender stereotypes. Here is some scientific articles that you won't read that explain why you are not correct.
http://www.livescience.com/226...
http://www.bustle.com/articles...
http://io9.gizmodo.com/5879647...
The problem is, that we tend to take the outlier as the rule, when trying to break norms. We should accept that norms are those for a reason, while not excluding those that are outside those norms. There will be girls who like trucks, and boys that like dolls. Blaming stereotypes solves nothing and doesn't actually progress understanding.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
Not only did I read them, I read them well enough to pick out the parts that contradict you.
"Wallen approaches the data more cautiously. "It's hard to interpret what the looking data mean because we don't know why people are attracted to specific things. Clearly children recognize that certain objects in their environment are appropriate for certain activities. They could be looking at a certain toy because it facilitates an activity they like," he said."
"And just because you play with a particular toy doesn't mean you only interact with it in a certain way. Dolls can be made to race, and trucks can be cuddled and carried around. Toy preference alone doesn't tell the whole story. So when it comes to girls, boys and the toys that want to play with, any generalizations are murky at best."
"Also, the researchers suggested a social rather than biological basis for the behavior. Because regular stick-carrying hasn't been reported in other wild chimpanzee communities, they proposed that that young Kanyawara chimpanzees may be learning the behavior from each other as a way of practicing for adult roles – a form of social tradition passed between juveniles previously described only in humans."
Oh no! It seems that the science isn't quite settled here! It seems that socialization DOES have an effect! It seems that while NATURE ISN'T NOTHING, NUTURE HAS MORE OF AN AFFECT THAN YOU GIVE IT CREDIT FOR. See, that's the thing I said before.
Keep it up, chum. Maybe one day you'll find someone that accepts links as proof without reading them better than you do.