Here’s what something reads like when you could have made a good point, but seem to have gotten lost in your own word salad.
How unsurprising, then, that the infamous 1984 commercial for the Apple Macintosh, which unleashed the personal computer revolution, featured a sexy, skimpily-clad woman shattering the gray political passivity of scores of lonely, propaganda-watching men.
Sexy? Skimpily-clad? What the hell is she talking about? It was just a woman in track clothes. She’s not sexualized at all. And what does any of that have to do with the rest of the article? Of course I could write that about nearly any paragraph in the appallingly poorly-written piece.
The Apple ad was supposed to depict someone athletic, vibrant, free from restraint. That she happened to have breasts was just due to the fact that she was a human female (and the fact that Apple very much wanted to appeal to women, as opposed to other computer companies).
Also, while it’s a myth that female voices are easier to understand, both men and women do find female voices more pleasant and less threatening than men’s.
So for most companies there’s no percentage in making the default AI voice male — you’ll displease far more of your customers.
I think I get more annoyed when an article could’ve been insightful and interesting and was not, than when an article seems to be crap and I read it and turns out that, yep, it’s crap.
So how could this piece have made some good points?
It could’ve examined why both men and women find women’s voices more pleasing. It could have done some analysis on why threatening AIs often have male voices (HAL in 2001, Agent Smith in The Matrix, Sonny in I, Robot, etc.). It could have examined how gender relations will change as both men and women use increasingly-sentient sex robots (and yes, women will use them too — perhaps more than men).
It could have made some mention of how genderless robots (BB-8, R2-D2, Jinx) are assumed to be male, and also why we humans need to assign everything a gender. It might have looked at how the Fembots in Austin Powers is a satire of and pushback against the idea of feminized embodied AI. It might’ve actually discussed the film Ex Machina and its huge relevance rather than just using some de-contextualized photo of Alicia Vikander as Ava, and also used that to segue into the fact that many men do actually want something with free will that nevertheless belongs to them. And then gone into the morality of creating any AI of any gender in the first place.
I could go on.
Article fail. Enough said.
Yes, I do think that writer was off when it came to the 1984 ad. Perhaps she read the woman as sexualized because she’s the only person with any individuality in clothing or action in that commercial? Everyone else in that commercial is deindividualized and desexualized.
The article linked to from the Salon article is slightly better. Although I don’t know why when comparing malevolent AIs & gender breakdown you wouldn’t mention GLAD-OS in addition to Ava. I also thought of Sushiborg Yukari It seems like the rise of fictional malevolent female AIs reflects an anxiety about labor, especially emotional labor and women that might not have been there in prior years. A lot of these lower level jobs have a heavy emotional labor component, and I suspect the whole “service robot rebels and program-murders humans” genre also reflects displaced anxiety over how we treat people who do that kind of work. The less genuine affect a service job allows for and the more stereotyped the responses expected from someone doing that job are, the more likely it’s a candidate for a service bot.
I think Cook was on the right track at first examining gendered labor in service professions as it relates to our expectations and actualizations of AI, but she probably does not have enough background in anything else to really do a deep analysis (true of most academics).
Apart from that, the writing is just very poor. It’s like something spewed by a freshman attempting to sound intelligent. Points are poorly-made and not well-supported. There is as much obfuscation as elucidation. And it’s wordy without being revealing.
“I suspect the whole ‘service robot rebels and program-murders humans’ genre also reflects displaced anxiety over how we treat people who do that kind of work.”
Agreed — that genre has seen some movement recently, and it reflects real anxieties.
About GLAD-OS, I don’t really play video games and am only vaguely familiar with that AI’s characteristics. I wouldn’t have the patience to play more than a few minutes of the actual game. Leaves a gap in my knowledge, for sure, but games and I just do not get along.