Okay. It's time for an AI rant.
My nephew is 13 years old. Whenever he writes a paper for school, I check it over and fix all of his mistakes for him. He said to me, "Maybe I'll proofread your paper for you in exchange," meaning one of the scholarly articles I write for work. I said, "Cool," and gave him the file. And he said, "Well, this is full of errors! See, you always say you have a lot to correct on my stuff, and look at all the stuff you got wrong!" And I said, surprised, "What? Where?" Because I'm sure there are typos in the draft I sent him, but not, like, that many.
And then he pointed to the screen and said, "Look at all the blue and red lines you have."
And I said, "Yeah, but those are wrong. Like, those are blue and red lines I'm ignoring because the computer is wrong." And then I paused and added, "You know you can't proofread a paper by just looking at the red and blue lines, right?" And he gave me the blankest look, because that clearly is EXACTLY what he thinks. And it became even clearer suddenly why, whenever I correct something on his paper, his immediate reaction is, "It didn't have a blue or red line."
There's a very good reason for that: THAT'S BECAUSE THE COMPUTER ISN'T SMART ENOUGH TO KNOW THAT IT WAS WRONG.
I am so tired of being sold the idea that computers are better than humans and so we should just outsource everything to them, which is clearly the lesson my nephew is absorbing in U.S. middle school. COMPUTERS ARE NOT BETTER THAN HUMANS. Like, maybe they are better at humans at crawling through rubble to find people trapped inside. They are also better at preserving things in a searchable format. Things like that. Very limited circumstances.
I don't want to sound alarmist but everything I hear about people using generative AI freaks me out. It's not just that I'm freaked out by people being like, "I use it to write novels!" (Although I don't see how they do, I have tried to have it write fiction for me and the output was truly terrible.) But I recognize my bias around creative writing and so no one needs to credit my views on artificial writing. But! Other things are alarming, too! "I use it to brainstorm x, y, or z." But...why? Why not just...use your own brain...to...brain...storm? The computer doesn't even have a brain to brainstorm with! And you might be like, "But it comes up with things that my brain would never think of!" So would other people! You could also brainstorm with other people! Or even through Google to see what other people have thought before you (not AI). Please don't belittle the wonder of thinking.
I just feel like the marketing around generative AI boils down to "Wouldn't it be easier not to use your own brain to think about things?" Everyone. No. It would not be. Please just trust me on this. I'm not just an old person who is out of touch with technology or something. I promise. USE YOUR BRAINS. IT WILL BE OKAY.
Yeah, this.
What scares me is seeing educators, from K-12 up through graduate programs, get on this bandwagon. I mean these are people who should know better and in many cases, I believe, they do. But the fear of being 'left behind' is strong, especially in the humanities where for decades people have valued what the humanities teach us less and less. So you see people using generative AI in the classroom, and in their own research, and a lot of the time what they're doing is something that could absolutely be done better *without* AI, but they feel pressure to show that they can Use the New Technology whether or not it is necessary or even helpful. To Stay Relevant or whatever.
I'm sure there are a lot of AI applications in fields where the ability to do pattern recognition on an unimaginably vast scale is actually helpful. But over and over we're discovering that AI does not and cannot know what anything *means.* This makes it utterly inappropriate for any task involving creativity, expression, or communication.
