As educators, we are faced with the fact that AI is becoming more and more accessible to our students. And, whether we like it our not, most of the education piece for our children and youth around AI is being left to us in the education field.
Something that you may have noticed about ChatGPT or any of the other Large Language Model (LLM) chat bots that are becoming available, is that they are being presented to the public in an anthropomorphic way. And this is on purpose. ChatGPT, and other LLM’s appears to “write” the response in the text box in front of you when it could simply just PLOP a page of data onto the screen. But when we see it “writing” in front of us it gives the sense and feel like this is a human typing out a response to us.
Not one person in the world looks at a PowerPoint presentation screen and says ” Wow, that PowerPoint software is really smart.” Instead we know that it was a human that put that information on the slide. But with AI, it get’s blurry for us, because it is a computer that is “writing” the information for us. If LLM’s simply plopped a page of information and text in front of us, we may tend to think – ‘Yawn. This is just a fancy search engine that plops me out results.’ But if we appear to witness the text being created in front of us, we tend to be a lot more engaged with it. Perhaps we even trust it more?
The interaction with the server is also designed to reflect a more conversational tone. Apparently old school DOS command lines telling a computer what to do just didn’t engage people deeply enough to want to spend hours and hours interacting with it.
dir c:*.csv /s /b > c:\users\todd\desktop\csvfiles.txt *
* I don’t even know what this does. Maybe look for CSV files on Todds desktop? BORING
AI design is about creating interactions with the user. To do that it is conversational. LLM’s like Pi.ai are actively promoting the fact that their chatbot is more polite and witty and fun than the others. ( Full disclosure, I think it is.) And witty and fun and engaging means that people will interact with it more freely because they get a sense that they are actually conversing with a human being rather than a server somewhere.
Here is the issue for us in education. AI chatbots are being created everyday to be more engaging and as a result, our young people are more inclined to get the sense that they are interacting with another human being. A friend. A buddy. But they are not. It becomes our job as educators (and parents) to change that narrative for our young people and help them to understand that AI is a tool. NOT A BUDDY. President Barack Obama talks a little about this in a recent podcast. 1
How do we do that? First off we explain what AI is to our students. We talk about the way that AI is constructed and what happens when we type in a request to a chatbot. We demonstrate some ways that AI can be helpful to us in the classroom. It can help us spark ideas for our writing. It can help direct us to places where we can learn about new things. We can talk about how it will affect our society. It might be better at reading medical imaging and work alongside doctors and nurses, but it can also replace many peoples jobs and what the effects of that will be on our communities. We need to be honest and open about AI to our young people. We need to teach them that AI is a tool they can use, but not a buddy.
Carroll Media 