When I graduated from high school in 1972, we had two calculators, one in the math lab, and one in the chemistry lab. They took up a couple of square feet on the tables and had a power cord the size of my wrist. They added, subtracted, multiplied, divided, and did square roots. By the time I headed to grad school in 1981, I had a $50 calculator that fit in my shirt pocket that could manage integrals (something I barely remember how to do).
At the end of my undergraduate studies, computer searching arrived. Instead of looking through Bio Abstracts, writing down lists of numbers that referred to articles containing a certain keyword, finding the union and/or intersection of those lists of numbers, looking up the abstracts for each of the numbers, determining which articles were relevant to what I was researching, then going to the Interlibrary Loan Desk to order reprints, I could “buy” the numbers for each keyword (I think it was $5 a term), then get the computer create the unions and intersections. Then I went to Bio Abstracts, looked up the numbers, and went to the IL desk.
Before long, computer searching actually returned not only the article, but the place in the article that was relevant to the keyword. No longer did the researcher need to actually read the whole article; they could extract the relevant part and move on. I remember being very concerned at the time that this would allow people to extract the relevant portions without understanding the context.
In grad school, I discovered computers. (Good thing, too, as I never finished my MS in Zoology and made my living instead as a “computer consultant.”) When I started using computers, I realized my thinking became much more reductionist and binary, as this is the best way to think about computers, which are, of course, binary. When the fledgling radio call-in show, “Talk of the Nation,” had a segment on computers in primary education, I called in. Thanks to the miracle (at the time) of one-button redial, I was able to get through and posed a question about how computers would change the way students thought. The guest’s response was interesting. “You are in good company. People were very concerned when books started becoming widely available that people were going to lose the ability to memorize.”
In the short time that AI chatbots have been around, we have seen them being incorporated into education. In my technical classes teaching HVAC technicians, I don’t require them to memorize things. Most everything they have had to memorize in the past is available to them on their phones. Having them memorize things today would be like insisting that they do all their calculations without a calculator. In fact, I really only have one learning objective: critical thinking.
But there is one fundamental difference between AI and a calculator. A calculator applies hard and fast rules of arithmetic. There are specific incontrovertible answers. AI, on the other hand, uses statistics to predict what will come next. Instead of “right” answers, there is the probability that an answer captures the question the user is asking. But unlike with good statistical analysis, there’s no estimate of variance. The chatbot returns a well-written (if bland) answer based on the probability that the words it puts together answer the question. Yes, it uses a huge data set to make the prediction, but as we have seen, chatbots can confidently return information that is not only misleading but factually incorrect.
Another concern is that as these AIs start to generate more of the information AIs are trained on, the less new thinking will be incorporated. I hope that as we learn how to use AI as a tool, it will serve to promote fresh thought. Just as calculators freed us from the mundane, freeing up time for larger thought, AI may do the same.
In the meantime, we will use our creative thinking to figure out how to incorporate this powerful, and sometimes fickle, tool into academia and our lives.
Postscript
This essay was written by me without the help of AI, with the exception of Grammarly, the AI grammar and spelling checker. When my college gave access to Grammarly to students and faculty, I was amazed to find that my grammar was less accurate than I thought. Since then, I have given over the part of my brain that was obsessed with grammar and spelling to Grammarly, freeing it up to become obsessed with other things. And I’ve watched as my spelling has gotten worse.
There are lots of things—like vehicle repair and maintenance—that I used to do myself that I’ve given up. I love the idea of being self-sufficient, but I early on shifted from wanting to do everything myself to wanting to be able to do those things. I probably have enough property to grow all my own food, but the time it would take would displace other things (like writing this essay) that I find more valuable.
Robin McClellan 