Let's Not Automatically Demonize ChatGPT

Recently we purchased a dishwasher. After a nice conversation with the salesperson, I waited for a date and time for delivery and installation, not realizing that Jenny– a chatbot– was in charge of that process. Jenny was great! Cheerful, responsive and efficient. The delivery people (real people, not chatbots) arrived right when she said they would.

This is not a plug for the appliance store or a review of the dishwasher. Rather I am making the point that AI is already a part of our day-to-day lives even though we may not realize it. Recently, there has been a lot of chatter about the beta version of Open AI’s ChatGPT.

Models like ChatGPT have implications for fields including business, commerce, health care and … education.?

Open AI’s website says:

We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.

Sounds promising. But I can assure you that the K-16 education sector will see it as a huge threat. Cheating gone wild! Students will ask AI to write their essays, do their math homework, write code for them, etc. Schools and colleges will convene task forces to create policies and procedures to insulate them from AI, and disciplinary committees will determine what punishments to mete out to students who engage with the technology.

How enlightened. Or …maybe we could start somewhere else? Some thoughts and observations.

• Technology does not cause cheating. Sure, it is a tool that can facilitate cheating but cheating has been around since the beginning of time. So as educators, let’s not think about ChatGPT as a cheating technology that we need to keep out of the hands of students. Which, by the way, is impossible. We should instead be spending our time helping students focus on how to best leverage technology for deeper learning. Why not have students engage with and evaluate ChatGPT and other forms of AI? Where is the potential? What are the pitfalls? Yup, AI makes mistakes. Because … humans design it. By the way, textbooks have always been full of mistakes. So let students dig in.

• How are we trying to shape learning? In our non-linear, figure it out world students need to immerse themselves in skills that are what I call the New Basics: Creative Problem Solving, Collaboration, Iteration, Tech & Media Literacy, Empathy, Visual Communication, Presentation Skills. You may have your own similar list; the most important thing to recognize about these skills is that they are not soft skills! In fact I would argue that content acquisition is the new soft skill. Content in itself should no longer be an end; it’s a tool. Content is easily accessible – ChatGPT is just one more resource – and is irrelevant if we don’t know how to apply it.
• How do we think about “academic integrity”? The answer is not fixed and has evolved over the years. Writing on a computer was not “real” writing. Collaboration used to be called cheating. Calculators were going to render students numerically illiterate, and the College Board took a long time to allow graphing calculators on the SAT … fearing students would… do well? Teachers would fail students for using Wikipedia in research. Will AI again push us to rethink the nature of academic integrity?
• Where is the potential? Do AI and ChatGPT have the capacity to make it easier for teachers to personalize and individualize learning? Could we eliminate textbooks (as an object and as a construct of knowledge dissemination)? Textbooks are all about one size fits all learning. And, by the way, they are biased, expensive and not exactly environmentally friendly. Can ChatGPT help create an IEP for each student?

This is a very new technology that has, appropriately, engendered healthy skepticism of all sorts.Will teachers feel the need to double down on in class sit down exams at the expense of more project based individualized work? Will colleges place greater emphasis on flawed standardized testing? Will schools back away from personalized learning as a goal? And as Boston Globe business/technology columnist Hiawatha Bray writes … What if ChatGPT isn’t actually that good???

But rather than demonize ChatGPT, let's encourage schools, teachers and students to dig in on its potential and pitfalls and in the process evaluate and challenge our assumptions about what learning and teaching need to be about.

As news sources say … this is a developing story. I’ll check back in a month or so.

Peter Hutton was Head of Beaver Country Day School for 28 Years and is about to launch The Project Lab - a platform for High School Students to work with industry experts to immerse themselves in real world problem solving and thus stand out in the college application process.