Here comes the AI apocalypse!
Fiction and films and such warned us of the dangers of artificial intelligence (AI). They told us how it would end the world. They were probably right. But it’s a pity they still somehow got it all wrong.
You see, in the movies or whatever, AI gains self-awareness and takes over the computer systems and dominates humanity or launches global nuclear war or something. You know, like The Terminator. But it’s so much worse in reality.
First, though, let me say that I’m not a luddite. I don’t hate technological innovation. I complain about a lot of things when they go wrong, but who doesn’t? I love the convenience of my cell phone; the constant robo-calls, not so much. The internet in general, for example—from social media to search engines—has been amazing. But it’s also morphed into some dark forms over the years. Same with AI. In science, for one thing, AI is great. If you need to develop a new drug, going through tens of thousands of subtly different chemical combinations and trying to predict likely results of each would take ages for humans manually doing it. But AI has been used plenty in recent years to take care of that step and narrow things down to a few best options that then humans can dig their hands and brains into to test and refine.
Like many seemingly good things, though—self-checkout being one of them—it doesn’t always turn out like anyone hoped, or it gets turned against the average person to the sole benefit of the powers that be, whether corporate or government or otherwise.
So, what’s my beef with the current state of AI and how it’s being twisted and misused? Oh, several things. Let’s tick them off.
First, AI has been increasingly employed to do things like facial recognition or even predicting behaviors, ostensibly to find criminals or even head off people before they commit harm of some sort. But here’s the thing: People act like technology isn’t biased, but if it’s programmed by humans and if it draws its information and conclusions from human assumptions and experiences, it becomes biased itself. AI struggles to differentiate nonwhite faces and to identify speech patterns of people who aren’t white. So, basically, the technology largely created by white people is focused on getting it right for white people and wrong for nonwhite people, especially Black people. Kind of like it’s always been in real life, only worse and more invasive. White people already tend to assume the worst of Black and brown people and interpret things like louder voices with anger when they don’t do the same with white people. Now the AI is doing it on a bigger scale while being assumed to be impartial.
Second, AI is moving away from things like analysis and now into “creation.” You’ve probably used this plenty yourself in the past year or two with TikTok art filters to change your appearance to something fanciful or in art generation. It’s generative AI and it’s now all over the place, with ChatGPT and the like, writing letters for you or writing whole entire research papers or creating pictures or writing stories. And, that’s awful. People aren’t putting the work into academic assignments, so they aren’t really learning anything or sharpening their thinking. Worse, the AI isn’t really doing anything original or actually creative. It isn’t proving that it’s better or just as good at writing or art or anything else. What it does is use computing power to quickly search the internet, pull together a bunch of things, steal bits from here or there and combine them, and make something that mostly makes sense but didn’t come out of any true thinking process. And this is hurting artists. People are using AI that steals from original human art instead of supporting artists. Artists of color or from other marginalized communities are particularly feeling the sting, as they often struggled before and increasingly are dead in the water now. How long before AI is cannibalizing so much of the same material and territory that everything looks and sounds exactly the same? And how long before all those artists are homeless or all working at shitty shift jobs because we’re using generative AI instead of paying for actual artists’ work and supporting what they are good at? And how long before most of the world is acting like the people in Idiocracy because they had AI do all their assignments at school for them?
Finally, AI is destroying our infrastructure and environment now. Because of the explosion in use of AI for generative AI instead of good old research purposes, companies have to build more and more processing centers for more and more intense computing. Remember those stories of how bitcoin mining (which uses computers and algorithms and calculations) were eating up tons of electricity and straining grids and creating waste all for the dream of cybercurrency which is mostly a pyramid scam but continues nonetheless? Well, now it’s generative AI adding to that. And so cities and states urge us not to use air conditioning as record heat hits us, or urging us to keep our A/C set so that our homes are in the high 70s or low 80s, because of strains on the electric grid. Then they turn around and welcome in AI companies that are draining record amounts of electricity to create fake art and soulless writing and more.
So, yeah, AI is probably going to kill us. Just not the way sci-fi told us it would, and not because the AI itself is trying to be malevolent. It’s just another corporate and government tool, one of the best yet, to kill our spirits, control us and waste precious resources all for a few more bucks.