Lin (2024) also wrote on what the general consensus is in my readings, and I agree with it, that AI has a long way to go before it can completely do the many things a teacher do in a busy classroom. Lin (2024) also pointed out that the significant limitation of AI tools, that they lack the interpersonal interaction and communication skills comparable to (human) teachers and students’ peers.

Harvard Graduate School of Education video “Experimenting with AI in the Classroom” featured a high school student and two researchers in AI talking about AI in education.

The student learned about ChatGPT by experimenting. He argued that ChatGPT can be used in good ways to help students with their work. He says that students use ChatGPT to cheat on busy work and homework. He makes an interesting point that good students could write their essay themselves with the time and effort it takes to get ChatGPT to write the essay for them. He said that students cheated before ChatGPT and if they want to cheat they would cheat without ChatGPT. ChatGPT makes it easier to cheat.

Dora Demszky, a researcher developing AI tools points that though a lot of focus is on how students are using AI tools such as ChatGPT and cheating, AI tools can be useful for teachers. I like that Dora points out that human teachers are pivotal in many aspects of learning such as motivation and role modelling and AI can never replace teachers in these areas. And I agree. The human element in teaching is important and the AI can never do that.

Also the technology is not there as yet. In some things that it can do, it can get you from zero to somewhere but it cannot get you all the way. Which is why teachers should always be in charge and the AI should be in the loop. It should never be the other way around. AI tools can help teachers in some areas of their work where they need help such as lesson planning and other classroom administration tasks.

Dora Demszky talked about an AI tool they have been developing, an AI coach. She notes that though it works, it does not work perfectly and there are always ways to make it more effective. We can always improve the technology. For me, she made a very important point, the technology can be improved but you can never get it perfect, just like you can never make a human coach perfect. And the same coach (human) may not work for everyone. Similarly, some users may hate the AI tool. So using AI tools is very personal. You can never make one size that fits all. For me this is a very important point when we think about AI tools such Intelligent Tutoring Systems (ITS) and Adaptive Learning Systems. The AI system may not work for every student and some student may hate this particular AI.

What Dora said made me remember something I read sometime ago. “I cannot be a perfect teacher if I am to prepare you for an imperfect world”.

Houman Harouni said that we have been focusing too much on students’ learning and students’ outcomes. Are students learning? What are they learning? How are they learning? Quoting Harouni, “Maybe the real criteria is not learning. It is not even work and performance. Maybe it is agency. Maybe it is about empowerment”.

Houmani said, take anything that you want to learn right now, you can find resources on it from the Internet. For example, if you want any book, you probably can download it from the Internet, and maybe for free. Houman Harouni then pointed out that now we have AI tools that can read the book for you and analyze it. I know about this because my students do it all the time when I assign readings to my students in my classroom. My students are able to use AI tools to write the report and make the presentation for the assignment I would give them to do.

Like Dora, Harouni is of the opinion that if the AI tool is not there yet, it is just a short distance from getting to the point where it can do it for us.

Quoting Harouni again, “The real question is, how do we have students, and, first of all teachers who feel empowered and feel like they have agency in relationship to the world in general and then in relationship to this technology? And you learn agency by understanding how you are limited, by going to the edges of your ability. That’s what experimenting is.”

Houmani describes experimenting with AI tools. “I asked this question, it did this. How much further can I do? What else can I ask? What kind of question can I ask? How can I push it to do something that maybe at first it could not do?”

Houmani is of the opinion that more needs to be done in getting teachers trained in using AI and get it know it well.

On the question of advice to teachers. Houmani returns to the point that maybe learning is no longer the goal and that have real implications for the role of the teacher. “On one level we say it is about student learning and student performance. But a lot of it was about students performing school – performing work for the teacher to get a grade, to get to the next level of school, to the next level of social life, etc.” “I don’t write because I am given an assignment, I write because I want to, because there is something in my mind that I want to explore, that I am trying to communicate …”

Houmani is of the opinion that teachers need to do something that is inherently human, go to some place where the AI cannot go – as yet.

References

Harvard Graduate School of Education. (2023, Sep 6). Experimenting with AI in the Classroom. [Video]. YouTube. https://www.youtube.com/live/qWJXMCEAgBY

Lin,Y. (2024). Adapting to the AI Era: Higher Education’s Opportunities and Challenges with ChatGPT. Lecture Notes in Education Psychology and Public Media, 40, 137-143.

Williamson, B.,  Molnar, A., & Boninger, F. (2024). Time for a Pause: Without Effective Public Oversight, AI in Schools Will Do More Harm Than Good. National Education Policy Center.