With the sudden increase in AI with the entrance of GPT3, faculty find themselves scrambling to arm themselves with good information about the educational ramifications of AI. Before looking at the implications for the classroom, learning about the basics of AI is imperative.

In the first session of a four-part series about AI at the Adams Center this semester, Dr. James Prather, Dr. Marisa Beard, Dr. Stephen Rektenwald, and Amos Gutierrez introduced the basics of AI.

AI is a broad term used to describe a family of related approaches to automated decision making. These can be based on things like rules, state space search, reinforcement, or pattern recognition.

There are three kinds of AI:

Supervised

  • Training on labeled data
  • Classification tasks
  • AlphaZero chess engine
  • Most NLP models

Unsupervised

  • Data analysis
  • Clustering
  • Finding hidden patterns in data
  • GPT-3/ChatGPT

Rule-based

  • Traversing a state space to find an answer
  • Simple games
  • Expert systems

What is ChatGPT?

  • Natural Language Processing (previously used for text translation, spam detection, sentiment analysis, speech recognition)
  • Large Language Model (175 billion nodes)
  • Can be trained on multiple datasets: art, programming, essays, poetry, etc.
  • Prediction model: given the input, what is the best output?
  • So, should we be worried?

There is no shortage of information and opinions about what AI will do to our education systems, teaching, and learning. The Adams Center created a significant document with some of this information here.

Here is what we know right now:

  • It’s very good at writing human-like text…
  • …but not so good that it’s without hints and signs. (for now)
  • Students will use it. You can’t ban it.
  • It has some great, legitimate uses!
    • Starting, brainstorming, making and then editing, quickly finishing boring tasks, making assignments, interpreting feedback
  • There are already tools for recognizing AI generated images and text.
  • Turnitin will be launching an AI generator tool this Spring.

As we continue through the AI series this semester, we will consider some of the following questions:

  • What about ethics considerations? Whose work is it?
  • How do we rethink learning objectives and assignments in light of tools like this?
    • How can we get students to do the work and think critically?
    • How can we utilize this tool and implement it into our curricula?

See you in the Adams Center for the rest of the AI series:

  • February 16: Ethical Considerations
  • March 9: Assignments That Use AI
  • April 12: Assignments That are Harder to Use AI