I.

When the new AI era came about almost three years ago, I was sure that it really wasn’t going to change much in terms of education. I think I am correct so far. The true bent of education is to get students to practice solving problems and think deeply. And also communicate effectively, work well with other humans. There’s lots of stuff that AI is not really going to change all too much, even when AI becomes digital water, so in-built in our work that we hardly notice we are using it.

But the thing is, the AI era in the classroom is exhausting, which is the opposite of what has been advertised. Take me for example: the layer of policing I now have to open myself to when reading student work is way too much for my regular mammalian brain. It’s pretty crappy when you get into a profession to inspire and then have to spend time policing, doing counter-espionage work to make sure that you are teaching content and skills as well as honesty and humility and the importance of failure.

For instance, I just found out about this company: Undetectable AI. From what I’ve seen, its sole purpose is to sell undetectable AI-written papers. With a $5 per month subscription, it will not only help you humanize and auto-type your essay (auto-typing would counteract an important way we check for AI in the Revision History of Google Docs), but it will also be a tool for good–it will help teachers check to see if student papers have AI. And it will also students confirm that their essays won’t show up as AI.

I didn’t find this by looking for it. I found this because a student’s paper had very suspicious and measured looking copy and pasting. It was word for word, and it was represented in Revision History as if the typing had a paced beats per minute. I wondered if someone made a fake typer for a Google Chrome extension and, voila, I found it.

Since I wanted to check to see how it worked, I made an account. But I couldn’t use it without entering my credit card for a seven day trial. So I ended that experiment. Though, I am no getting some pretty interesting emails from Undetectable AI. You can see who they are targeting:

I have caught many students writing with AI, despite the fact that I have run my class around normalizing failure as part of the process. But the reasons for using AI are various, probably more so than the usual plagiarism of yore.

Certainly expectations need to be changed. They had to when plagiarism became easier with the advent of digital copying and pasting.

Whether we handle plagiarism well today or not, we are in the business of teaching, which also is synonymous with reform. I don’t think school rules need to change back to the strictness of 40 years ago: you plagiarize, you reap the largest consequences such as expulsion or a permanent mark in the record. I’m not sure it’s helpful to mark a youth’s mistakes for life, but it shows desperation that such consequences seem tempting to even me.

Barring in-class writing or oral examinations–both of which don’t prize process, which is the gold standard of learning–there is not much teachers can do to stem the flow of AI. We do more police and espionage-like work than ever. Like in the crime world, there is simple crime and smart crime–those who know how to get away with it and those who are just desperate.

II.

I was very fortunate that my parents gave my pre-eighteen self an international life. I bring this up only because of those passport control people you have to go through in order to “arrive” in your country of origin or visitation. I am not trying to knock people that work in that area where you get your passport checked, but young me wondered what kind of person got that sort of job. It seemed a serious duty to do for your country, but at the same time, it seemed like a stressful job, always on the hunt for liars that could turn out to be wanted drug-dealers or violent terrorists.

Later, I wondered what it did to them.

Meeting new people is already pretty hard. When we’re growing up, we are thinking about our place in the various social standings we are built to acknowledge. We want to be a part of a group. And whether we acknowledge it or not, we want to be accepted by everybody.

Who does a passport control person accept? Obviously they accept citizens of their country. And maybe they really accept citizens who have a criminal record–we’ve been waiting for you! But you have to be, by nature and purpose of the job, looking for those who you do not accept. It’s why there are passport checks. And, due to the fact that this is against our nature and is probably kind of a bummer to live in that sort of headspace, you need to practice some active dehumanization. It has to be quite impossible to empathize with everybody trying to get into the country. Humans can’t handle all of those emotions.

Now, let me clear something up. Dehumanization is mostly bad, but it’s also necessary. Take surgery: you must distance yourself from the person you are operating so as to focus on your job, which I’m sure makes the surgeon mindset much like that of a car mechanic or computer technician. Also, success is, by nature of the job, not always guaranteed. That is a lot of emotional baggage to deal with.

Furthermore, if we empathized with everyone in the world right now, in all their states of life, it would be a hodgepodge of emotion that would deter us from living our own lives.

As a teacher of 160 students, dehumanization is already part of the job. I need to be responsive to my students, especially when they are down, but I also have to be ready to move on. It’s impossible for me, time-wise, to do everything in my power to make my students feel the right way about their work in my class, especially in terms of feedback. So each teacher has come up with what seems to work with their class and their own personalities. This is quite silly to point out, since school should be the humanizing place, but it’s part of all our jobs in some degree. Especially when we get home–we have to distance ourselves and focus on our home life. It’s why working from home during the pandemic was so painful for some.

III.

The most human part of education is the process of learning, this area where all things can happen and all things are important to pay attention to. When students write, they think through what they know and make connections to new insights. It is the same process that occurs when we design anything.

With AI in the mix, that process of learning has become increasingly endangered. Already many students have a school experience that they feel is not tailored to what they deem as relevant. They have busy lives, and many of them take seven classes a semester. If my class sizes were smaller, and I could really personalize the learning journeys of my students, I could combat this lost-in-the-pack sentiment. But education is what it is right now, and the way it’s been going, I’m sure there are already classrooms in which dehumanization has reached its peak: students using AI to create work and teachers using AI to give feedback on said work.

I am in the business of humanizing education, so I am in a passport control mindset. I want my students to value their writing process. To do that, I need to show them that I am here for them.

When you add more passport control or policing into a profession that is supposed to be molding better humans, you have a lot of emotional swings. I was not built to do the important job of policing. And I think that’s what’s going on right now: some serious emotional back-and-forths.

The general teacher response to all of this extra policing is not quite like the charming way the famous Hayao Miyazaki (though not charming to the person he was talking to) responds to AI:

I am utterly disgusted. If you really want to make creepy stuff, you can go ahead and do it, but I would never wish to incorporate this technology into my work at all. I strongly feel that this is an insult to life itself.

What I’ve realized is that AI is much like any new technology related to information. Furthermore, it’s something we’ve dealt with ever since humans started telling stories. The proper use of information does not come blindly but with much scrutiny. AI will yet become increasingly good for general knowledge and large data calculations, but suspect for specific answers. It will be, of course, an algorithm to the end of its days.

For instance, two weeks ago, ChatGPT lied to me about a question I had about a book I’m currently teaching: The Catcher in the Rye.

It’s my first year teaching it, and I think my third time reading it, though I remember none of the first two times. I was curious, as I went through it, what movies Salinger made up and what movies were real. I asked ChatGPT this question, and it mentioned The Great Gatsby. In its explanation, it acknowledged that the first Gatsby movie came out in 1949 and that Holden is not actually referring to the movie but the novel. Little did I know, I was two pages before the mention, but I was curious where it was mentioned. So I asked it:

The lying (or the preferential word, denoting a lack of intention, “hallucinating”) is prevalent in all the models, and I have seen it a bunch in both Claude and ChatGPT.

We must think of AI like we do any internet search tech or nonfiction book: AI can point you in a direction, like GPS, but that is it. AI lives on a different plane of existence. It is the human that steers toward a destination, experiencing whatever the journey entails, that is most important. For when we arrive at our destinations, all GPS knows is that we got to a calculated point, not that we are visiting our friends our family, the conversations that are spawned from our meetings, the connections we feel from togetherness.

It is already moot to separate those who use AI from those who do not. We all use AI: if you are typing in a search bar, you are most likely using AI. The dastardly thing is what AI will do to our understanding of process. But if we value only the product, perhaps my worries are all moot. Then again, if we only value the product, education needs to change extensively regardless.

Things have changed, and yet the yin and yang grappling in education are the same. We are yet at the mercy of efficiency versus humanization. Do we spend more resources on people or tech? This is something education cannot solve for itself. It is community-wide, for this is the nature of the system we have built.

In the end, humans must humanize.