Ask An AI - ChatGPT’s advice on how to teach and test in the AI era.

2 minute read

Author: Kevin Geiger

Original Source

Poster

The challenge of how to address AI in educational venues (as a tool? as a threat?) and how to integrate and/or regulate is a hot topic of discussion at schools and universities worldwide.

As a sometime professor with a vested interest, it’s clear that an educational evolution — towards AI-assisted education and project-based learning for pedagogy and snap quizzes and oral exams for evaluation — will be required.

But that’s just me. Why not go to the well and ask the bot what it thinks?

So, I prompted ChatGPT…

“How can schools and universities evolve their teaching and testing methodologies to address students’ easy access to AI tools?”

ChatGPT replied…

Schools and universities can evolve their teaching and testing methodologies in light of students’ easy access to AI tools in the following ways:

  1. Shift Focus: From memorization to problem-solving, critical thinking, and creativity. AI is great at recalling facts, but humans excel at nuanced thinking and creativity.

  2. Open-Book Exams: Embrace open-book and open-note tests, which allow students to use resources, including AI, but test them on higher-level skills, like analysis or application.

  3. Project-Based Learning: Engage students in real-world projects that require collaboration, creativity, and problem solving. AI can be used as a tool rather than a crutch in these situations.

  4. Authentic Assessments: Use assessments that mimic real-world challenges. For instance, students could be asked to tackle a current event topic, using AI to gather data, and then critically analyze and present findings.

  5. Teach Digital Literacy: Educate students about the strengths and limitations of AI. This includes recognizing biased information, fact-checking, and understanding how algorithms work.

  6. Oral Exams: These can be an effective way to gauge a student’s understanding and ensure they can’t simply rely on AI tools for answers.

  7. Collaborative Assignments: Encourage team-based projects where students have assigned roles. This can minimize the chances of one student using AI for the entire assignment.

  8. Regular Quizzes: Short, frequent quizzes can help ensure continuous learning and reduce the likelihood of students leaning heavily on AI for last-minute cramming.

  9. Incorporate AI in Curriculum: Integrate AI into lessons where appropriate. For example, teach students to use AI for data analysis in a science class or to check grammar in a language class, but also make them critically evaluate the outputs.

  10. Honor Codes & Ethical Training: Strengthen academic integrity guidelines to address the use of AI tools. Make students aware of what constitutes cheating in the age of AI.

  11. Time-Constrained Exams: When students have limited time to answer questions, they’re less likely to have sufficient time to consult AI tools effectively.

  12. Randomized Question Banks: Use technology to generate tests where each student receives a slightly different set of questions, reducing the effectiveness of sharing answers.

  13. Practical Demonstrations: In courses like chemistry or physics, practical demonstrations and labs can’t be easily replicated by AI.

  14. Personal Reflections: Ask students to submit reflections or opinions on topics, which are subjective and can’t be generated authentically by AI.

By evolving methodologies to focus on deeper understanding, application, and critical thinking, educational institutions can ensure that students derive genuine value from their learning experiences, even in an age where AI is easily accessible.

Not bad for a bot, eh?

I recall how in high school the most feared tests were “calculators allowed” or “open book,” so I surmise the same will hold true going forward for “AI permitted.”

Ultimately, the upshot of AI is that it is prompting conversations in every field — art, education, finance, medicine — about the essence of the endeavor.

FINALLY.