I read a Wall Street Journal column recently about new research about how AI does what it does. Researchers have come to the conclusion that AI is not as smart as proponents give it credit for. Apparently all it is doing is memorizing a massive amount of information, identifying patterns through trial-and-error, and then spitting it out in response to prompts. The AI isn’t reasoning in a human sort of way, but merely regurgitating. It occurs to me that too often we treat college students as a human AI as if an education is just about the memorization of massive amounts of information. One of our USF undergraduates referred to as the information dumpster. But a college education should not be about cramming students’ heads full of content. College students need to think not memorize, something that AI unable to do.
The Training of AI
AI is trained by inputting a massive amount of data using machine learning to synthesize and learn patterns. An AI has the entire world wide web as its source of information. Take any topic, such as employee job satisfaction and AI has at its virtual fingertips academic research, thought pieces, content on commercial websites, fictional accounts, and everything else out there on the topic. Ask it a question, like “what is job satisfaction, and it can synthesize all the definitions to create something that might be on target, or might be a distortion–a hallucination. There is no judgment or reasoning that can serve as a quality check. Over time, the training has become better so that accuracy has improved, but the AI is not thinking in the human sense.
We Are Drowning in Content
When I was a graduate student in the 1970s, there was comparatively little content being produced in my field. An I-O psychologist like me would be fine just following two journals, which combined produced 10 issues per year. More journals existed, but between Journal of Applied Psychology and Personnel Psychology you could be well informed about important developments. Today there are more than a dozen top journals and Journal of Applied Psychology has more than doubled in size and comes out twice as often. It is not possible to keep up with everything. The same is true in all fields, so the idea that you can learn a field and keep up is unrealistic. Human memory does not have the capacity.
College Students Need to Think Not Memorize
We have reached a point of content saturation that it is not possible to know everything, even within a narrow field. What is means to be well-educated has changed. College students, whether at the undergraduate or graduate level need to learn skills more than knowledge or facts. Yes, there is content that should be learned. Students need to understand their fields, which means concepts and major findings. But the emphasis needs to shift from memorizing content to learning skills in using content. In other words students need to think not memorize.
When I was an student courses consisted of lots of content that we memorized and spit back on exams. My upper-level undergraduate psychology courses focused largely on important studies, organized by topic. I would memorize details of maybe 100 studies in a particular class and demonstrate my knowledge on multiple choice exams. When it came to my doctoral-level comprehensive exam, there were hundreds of pages of content for me to memorize. A former colleague and pioneer in the cognitive neuropsychology field, Manny Donchin, used to say that the content he learned in school was unimportant because it soon became obsolete. What was critical was the methodology skills that he used throughout his career. In other words, college students need to think, not memorize.
Faculty should design their courses around a set of skills that they wish to help students develop. Critical thinking is at the top of the list. The content in courses provides a context and vehicle to teach critical thinking. Rather than memorizing content, students can be given activities designed to develop cognitive skills that they will use after graduation. Some specific activities include.
- Critique rather than memorize: Students can be asked to read a study and then critique it. What did the researchers do right and what did they do wrong? What are the logical flaws in their arguments? How could the study be improved?
- Debates: Students can be asked to debate issues relevant to the course. I used to have groups of students in a seminar read a point-counterpoint exchange, and then each group would have to role-play one of the positions as if they were the authors.
- Research projects: Students can be given a research project to do. I like to have students choose a construct and then write a few items to assess it. Then we collect data and each student conducts simple analyses and writes up results. Those results can be presented in class. In large classes, students can work in groups.
- Presentations: Instead of memorizing details of journal articles, students can be asked to give short presentations on those articles. These can be general overviews, or they could be more targeted where they are asked to answer specific questions.
- Fact Checking: You can ask students to take an article and fact check how citations are used. If the author cites something in support of their position, can you find the evidence in the cited source that the author is claiming? This has become particularly important in the age of AI where faculty ask students to use AI to answer a question and then fact check that answer.
There are many other activities that can be used to enhance skills. What is used depends on the nature of the course and the objectives. Emphasizing skills is becoming increasingly important. After all, college students need to think not memorize because after graduation what they can do will be more important than the arcane knowledge they learned in college.
Image generated by DALL-E 3. Prompts “picture of a college classroom with a professor with a fire hose shooting knowledge to the class” “make the wand a firehose” “try with a smaller hose and different looking professor” “the classroom is good. Have the professor look older and have them holding the hose with the knowledge streaming out. Do not show the hydrant.” “this is perfect. Just have the students younger and not all men”
SUBSCRIBE TO PAUL’S BLOG: Enter your e-mail and click SUBSCRIBE