AI in Writing: I Can Fix This!

I started this blog so I could ramble on about my sports-related opinions. But sometimes, I feel the irrepressible urge to ramble about other things. This is one of those times. And this time, I had so much to write that my original essay got a little away from me. On the recommendation of someone who had to read the original essay, I’ve decided to break this up into a more digestible four-part series. This is Part 4. You can find Part 1 here, Part 2 here, and Part 3 here.

Photo by Steve A Johnson on Unsplash

I Can Fix This!

Several months ago, I volunteered to give a few short lectures about computer vision specifically, and AI more generally, to a handful of San Francisco high school computer classes. The schools I volunteered at are generally considered to be “rough”, which generally translates to “poorer and more racially diverse”. I was warned prior to the lectures — the material for which was developed by the organization I volunteered through — that while the goal was to get the students to engage with the material, it would likely be quite difficult.

And you know what? It was. These students were thoroughly unimpressed with my technical background, they were even less impressed with my education background (can’t blame them on that one, I went to the “good” public high school in SF and I can’t say our social reputation among the other schools is the best), and I couldn’t even win these kids over with my “I love Bay Area sports” pitch. And I feel like they were giving me a comparatively welcoming treatment — during one lecture, my super-tech-bro co-presenter proudly bragged about how he used AI image generation so he ‘could be an artist too’, and you could have heard a pin drop in the room.

Ironically, that’s the moment I realized that I actually did know how to get these kids engaged. I grew up in San Francisco during the tech boom. I used to be exactly like them. When I was their age, I also hated the tech bro culture that was ruining my city — infecting the city with their hustle culture bullshit, raising house prices and rent, pretending to espouse liberal values while pushing regressive policies, all while preaching a philosophy that placed more value on technology and money than on creativity and humanity.

So after I asked them what the pros of AI and computer vision were, I asked them what the cons were.

Practically every kid had an answer.

You’re probably thinking, ‘what does that have to do with the AI in writing problem’? Surely, if I’m going to use several thousand words to complain about the effects of AI on writing, then I should be getting to the part where I tell you how to fix it, right? That’s the name of this section, after all. Why the hell am I suddenly talking about high schoolers?

First, let’s be clear: I don’t have a solution. Not really. But listening to those high schoolers think and talk so deeply about the pros and cons of AI — including the issues that come with using generative AI while writing — I felt hope. The next generation is more than capable of thinking critically about this new technology. It’s our job to help them.

When I help students who have been falsely accused of using AI in their essays, what concerns me most is that those responsible for teaching the next generation critical thinking skills seem to refuse to think critically about the AI tools around them. Of course, I don’t mean every school and every teacher — I know for a fact that there are many who have put in a lot of work to educate themselves and pass that education along to their students. One professor told me of a paper she graded once that included a clearly AI-generated, nonsensical figure with hallucinated values and words not found in any dictionary, English or otherwise. She now uses it to demonstrate to her students that using generative AI will harm their grade more than it will help it. Other professors make a point of explaining to their students that a base knowledge of a subject is necessary in order for generative AI to even be a little helpful.

But some professors and school administrators are so scared of the possibility that students will outsource their college experience to LLMs that they’ve banned them outright. In doing that, I argue that they are doing a great disservice to their students. Yes, of course students should be doing their own thinking and learning, and yes, of course students shouldn’t be using AI to complete their assignments. But to ban the use of these tools completely is to leave students unprepared for professional life. Employers, for better or for worse, are pushing the use of AI tools on employees. Since employers aren’t thinking critically about how these tools are applied (although they really should), the best thing a school can do to prepare its students is to teach them how to use these tools responsibly. Students need to be taught the limitations of AI, and they also need to be taught how to get the most out of them. As with so many things in life, education is key.

And while I believe a subject like writing should be among those that uses AI less frequently (as I hope I’ve demonstrated), I also can’t deny that generative AI has been useful to me in the writing process. Unfortunately for the witch hunting anti-AI purists (who ironically, use AI when accusing their victims of wrongdoing), there are some legitimate, useful applications of AI in writing.

Becoming a Luddite and completely shunning the use of AI in any situation is simply no longer an option. Fully embracing it and outsourcing all aspects of human creativity isn’t the solution either. On the subject of AI, I’m with Dr. Fei-Fei Li — the rhetoric around AI is far too black and white. It is a tool, and like all tools, the most important thing is how we — the humans — use it.

Have I rambled on for several thousand words in order to come up with an ‘it’s nuanced’ answer? Yes, I have. I can’t offer a grand solution; I’m not an expert; I’m not writing policy; hell, this essay probably won’t even change anyone’s mind. The best I’ve come up with is that we, as humans, have to think critically about how we use AI tools. Revolutionary, I know. It’s not like the “godmother of AI” wrote a whole book about how humans need to be at the center of everything AI. Along with building AI so that it has the benefit of humans in mind, I believe that a human-centered approach to AI also asks us to develop a baseline level of AI literacy. We don’t need to know exactly how these new AI tools work, but we do need to know how and when to use it responsibly; how we can use it for good; when to question its outputs and when to trust it; and to recognize its limitations. As people living in a new world that won’t shut the fuck up about AI, we owe it to ourselves.

More than that, we owe it to the next generation. Based on a small sample size of three high school classes, the next generation wants to think critically about AI. They know they’re going to have to use it, and they want to use it responsibly. It’s up to us to help them do that.

Also, stop using AI to accuse people of writing with AI. It’s fucking ridiculous.