Mind Readings: AI Content Detectors Deep Dive Part 2

Welcome back to part two on our AI and academia and AI detectors series. In this part, let's talk about why would a student or anyone use generative AI to do their work? It is human nature, and it is actually just nature, period, to use as little energy as possible to accomplish a task. That is all nature, from the smallest of animals to the most complex person. We want to do as as little as possible to conserve energy.

That is a survival instinct. What is the point of writing a paper? Especially if a tool can do a better job than human. Well, the point of a term paper is to have a student, this in theory is to have a student do research, think, consolidate their thoughts, and then express their thoughts in a in a coherent manner that shows that they can think. That is the point.

Is writing a term paper the best way to do that? Writing is thinking. Getting someone to write is a is a way to help them think. I write, I write a ton. My weekly newsletter that I write every Sunday.

I hand write, well, not handwriting, I type, but I write without the assistance of AI because an awful lot of the time I have ideas in my head that I want to consolidate, and writing is the best way for me to order those thoughts. It is not the only way. If you were to go back to ancient Greece, it's Socratic debate and Socratic and the Socratic method of talking through ideas. Writing was was uh a precious luxury. Sitting around and talking about something was the way most people did their thinking to talk something through.

We we still have that in our vernacular today, you know, talking something through. In a classroom, in a classroom setting, can you have someone do Socratic debate? Absolutely, you can. You could do it in triads, you could do the outer ring inner ring, you could do any of the methods of that system. And in doing so, particularly if students were not permitted the use of their devices, um, you could have them uh demonstrate that they could think about a topic.

It's asking challenging, open-ended questions. So we use the example from part one on uh how does the Supreme Court uh impact the economy? You could have literal debate. One side has assigned the perspective of yeah, here's how it impacts the economy positively and negatively, another side saying it has no impact whatsoever. And they can debate, they can prove try to prove their points.

The downside for institutional education is that doesn't scale well. It is very difficult to automate that. And if you read uh the very excellent uh the underground history of American education by John Taylor Gatto, uh you will recognize very quickly that the education system that we mostly operate with today in the 21st century, was optimized in the early 20th century by Carnegie and Mellon and Rockefeller. And the intent of the education system, at least within the United States, was to generate obedient workers. Uh, the education system was designed to generate obedient workers.

We have batches of product that we call grades, grade seven, grade eight, sophomore, junior, senior. We have quality assurance testing to ensure that a minimally functional product called standardized testing. We encourage conformity of thought, depending on your perspective, certain things may be indoctrinated into students. Things like, for example, in the United States, textbooks are rejected by some states for not having enough execu enough or sufficient examples of American exceptionalism, uh indoctrinating people into uh to a form of nationalism. And whether that's a good idea or not, that's up to the school boards in the country.

But that is what the education system is designed for. It's designed to create obedient workers. It is it and it worked very well in the early 20th century. We are not in the early 20th century anymore. It's a hundred years later.

So the question now is does the current education system serve students and provide good outputs for them to uh succeed in today's world? And that is a whole other topic for debate. However, the term paper, which is our focus and the use of generative AI, the term paper is a form of writing. However, most generative AI tools can do a better job than students can at writing on the same topic, sufficiently well prompted. And so a better question is how would you use generative AI to do a better job than just having someone spit out and regurgitate knowledge?

Uh Dr. Lisa Palmer wrote on LinkedIn not too long ago, and I thought it was a very useful piece saying instead of trying to catch student cheatings, have them turn in their entire threaded conversation with AI. Think about it as showing your work like you would have math problems, then assess their critical thought process. Did they build thoughtful prompts? Were they demonstrating curiosity or through the correct collaborating with their AI colleague?

Partial credit with thoughtful feedback from their instructors will create a much deeper learning experience than correcting a student who struggles with grammar. This is a good take. This is a good take because the process that you need to use to get great output out of gender of AI is thinking. You have to sit there and think. Now, if you just write a prompt like write in justification for the Supreme Court and how it impacts the economy, yeah, that's a lazy prompt.

That is that shows no research, that shows no thinking. And you can grade that. You can grade a prompt or a methodology. And you can say, Yeah, you know what? You really did as little as possible.

In part four, we're gonna actually go through the process of how you would how you would build a modern term paper, a term paper of the future, if you will. But for right now, I think it's worth saying that the process of using generative AI itself demonstrates thinking as to how well you prepare, how well you know the tool itself, how well you know how to write prompts, how well you know how to get good output out of the computer and understand what good output does and does not look like. At Framingham State University, which is a university near my house, the the senior class in uh the senior seminar in psychology, the the psychology department chair has students intentionally write papers with Chat GPT, and then they critique the papers that ChatGPT spits out. They say, what did it get wrong? What did it overlook?

What is factually incorrect? And again, this encourages students not only to be critical thinkers in general, but especially to be critical thinkers of the machines and say, hey, machine, here's what you got wrong. It's a very, very useful uh set of exercises. So wrapping up part two. The point of writing term papers is to help a student learn to think and to then for them demonstrate their thinking.

And there are other ways of doing that. The term paper is a relic of an older education system, uh, an industrialized education system for an economy that doesn't exist anymore. And if we want to advance our students and get them to be capable workers in the modern economy, instead of banning AI, we should be encouraging them to use it to grade their efforts with AI as a partner, um, grading their prompts, grading their thinking, their depth of research, their original points of view, not whether or not they physically typed out the words on on paper. That's uh going to do it for part two. In the next part, we're gonna put some AI detectors to the test.

So stay tuned for that. If you enjoyed this video, please hit the like button. Subscribe to my channel if you haven't already, and if you want to know when new videos are available, hit the bell button to be notified as soon as new content is live.