Mind Readings: How Should Educators Think About Generative AI?

In today's episode, let's talk about how educators should be thinking about generative AI. I got an email from my son's school saying, we have forbidden the use of uh generative AI tools like ChatGPT and our uh anti-plagiarism systems will be rigorously looking for uh uses of generative AI and we'll uh and so on and so forth. They went on for a bit. And I'm like, oh, they don't. Um I happen to know a decent number of the plagiarism tools, and they are really bad at detecting uh A, they're not great at detecting plagiarism, and B, they have very little chance of detecting well-constructed generated text.

Now, if you just write a one-sentence prompt and you copy and paste from that into a text document, yeah. But you also don't need software to detect that that was written by ChatGPT, because that always sounds the same out of the box without a lot of tweaking. So, no. Good try, educators. Here's the the thing that educators need to be thinking about.

These tools are not going away. These this is not a fad. This is not uh, you know, something like uh uh a certain cryptocurrency or whatever. These tools are not going away. And the reason they're not going away is because they enable better, faster, cheaper, right?

You can create content that is better than what some people are capable of, faster by a large margin, and cheaper. Anything that creates better, faster, cheaper is not going away in the world. That's just the way the world works, right? Because everybody wants better, faster, cheaper. Businesses are adopting these tools like crazy, as fast as they can, because again, better, faster, cheaper.

So education and educators need to have a real hard think about the things that occur within the classroom, the activities that we do, the exercises, what we teach students, how we teach students, how what learning looks like, and say what is the value that we provide? A lot of education is still stuck in an in a 20th century mindset that education is the gatekeeper, right? These are the gatekeepers of knowledge, and we'll impart knowledge upon the students. That hasn't been the case for a while, but it's definitely has not been the case since 2007, because these devices, which are in every student's hands, are the gateways to information now. Education and the classroom is not where you get information.

So telling students, oh, you can't use this tool that exists that literally the whole planet's using, uh, and the businesses are desperate to find talent for and paying $330,000 a year to find. We're not going to allow that in schools. It's like the schools that tried to ban the internet. Yeah, that worked out real well, didn't it? So, what is the role of generative AI?

How should educators be thinking about it? Let's take the standard term paper, right? Take a term paper. What does writing a term paper teach you? What is it?

What is it good for? Synthesizing data, synthesizing information, coming up with a perspective, perhaps. But ChatGPT can write a term paper, probably better than you can on any given topic. So what are you really learning? The mindset that education needs to adopt and rapidly is not being a gatekeeper of information, but teaching the skills of how to critically evaluate information, how to look at information and go, that's wrong, or that requires more research, or that has some nuance to it that is not explained clearly here.

And equipping equipping students with those capabilities. And maybe a term paper is that tool, but probably not if a machine can do it. What can't a machine do? What can a machine realistically not do in a classroom setting? Think about things like debate, rhetoric, argument.

Think about building points of view that are opinion based on data. Is climate change real? Well, there's evidence that there is. How would you build a case for or against it? What kind of critical thinking would you adopt?

And can you point out the logical flaws in any given position's argument? The danger that tools like ChatGPT pose are not cheating on exams and stuff, because again, let's be honest, the exams in their current format are not really all that effective. The danger that they pose to great the greater society is they speak very authoritatively, even when they are authoritatively wrong, right? Large language models write in such a way that comes across as very authoritative. And if you don't think critically, if you can't think critically, you read the app and go, oh, that sounds good.

I'll believe that. Even if it's totally wrong. Totally wrong. It is that subject matter expertise and the ability to think critically and look at a piece of information and evaluate and go, is that right? How would I know if that wasn't right?

What would I need to do to prove or disprove that? Those are the skills that 21st century Denmark or Sweden or Tuvalu or Malaysia or America need. Because we have a lot of people and a lot of outright hostile governments around the world using these tools to create misinformation and disinformation. How do you teach people to think critically? How do you teach people to look at a post on social media and go, you know?

A, I think that was written by a machine. And B, I'm pretty sure that's wrong. I'm pretty sure that's that's propaganda, not fact. That's what education should be teaching. That's how educators should be thinking about generative tools.

I had a conversation with Dr. Nicole Rossi at Framingham State University not too long ago, who was explaining that in the senior-level psychology class, they use Chat GPT right in the classroom. They will have it write about a specific topic, like you know, uh abnormal psychology, and have it put together a document, and then the student's role is to critique it, to say this is what it got wrong. This is the nuance that wasn't in there, this is what the machine didn't know or the prompt wasn't good enough to explain. That's how you use these tools.

You use these tools to demonstrate their capabilities, you use these tools to take care of tasks that frankly are not worth doing, like writing term papers, and most importantly, you use these tools as foils for helping explore students' knowledge. You write you have it write a position piece on the importance of recycling, and then you have students go research that to figure out: did it write something that was correct or did it write something that was factually wrong and that process of proving or disproving is the essence of critical thinking right think about every political or social issue that is a hot button issue how many people who are partisans of that issue have ever thought critically about the opposing point of view not many how many people could be taught to think that way if they wanted to everyone how many times have you looked at an issue that you're passionate about and say well if I was if I was of the opposing point of view what proof would I have that that point of view is valid sometimes there isn't right someone someone believing that the world is flat i it's not um and there really isn't any good way to disprove that point of view you can look at how people present that argument but it's you can disprove it with a your own homemade weather balloon a piece of garlic bread and a GoPro right which somebody did someone to uh attached some garlic bread to a GoPro and sent it up into space uh came back down they they tasted the garlic bread it was pretty frozen um but the footage very clearly showed that the planet was a big sphere like object right because it went so high up that you could very easily see a good chunk of the planet. That's the role of generative AI. Not to ban it from school from educational institutions, because that's like banning search engines. It's like banning word processors, banning calculators.

The reality is these tools are going to be sticking around, and students need to know how to use them. Sooner the better. Teaching students how to use them properly, teaching students how to write prompts, teaching students to QA the output. That's the value. And in doing so, you will resurrect a lot of those critical thinking skills that our current education system, particularly in the USA, where I'm based, is lacking.

The USA education system today is a manufacturing system. It manufactures workers, right? It was built by Carnegie and Mellon and Rockefeller in the 1930s and the 1920s in America to make factory workers. Think about it. What are grades of school?

They're batches of students. What is standardized testing? It's QA testing to make sure that the batches of robots that you're churning out aren't defective. That's what education is. That's not what the world needs right now, because we have actual robots for that now.

That's not where value is created. Value is created now in today's world with creativity for being able to create something that a machine has not created, for being able to look at an enormous amount of data and derive real insights from it, to critically think and find weaknesses in a competitor's strategy. All the skills that the world really values are not taught in schools, not in a manufacturing based education system. So for educators, look at what you're doing, look at the skills that are needed in today's world, and ask yourself: does banning an AI tool that speeds up the manufacturing process really detract from education? It shouldn't.

And if it does, you're not teaching the right things. That's all for this episode. Talk to you next time. If you like this video, go ahead and hit that subscribe button.