There have always been plenty of college students who preferred cheating to studying and doing their own work. That old problem has recently become more acute with the rise of artificial intelligence sites that produce passably good papers at the push of a button.
In today’s Martin Center article, Graham Hillard looks at the efforts to counter such cheating. He isn’t optimistic, writing, “On April 4 of this year, the academic-services firm Turnitin activated a software designed to catch a certain kind of student plagiarist. As has been widely discussed on the Martin Center’s website and elsewhere, the next frontier in academic dishonesty is ‘student’ work that is actually machine-generated — by ChatGPT, Gippr AI, or one of the other artificially intelligent ‘bots’ that will doubtless follow. Turnitin claims that it can now detect this artificial prose with 98-percent accuracy. Yet the most likely development may well be a machine-learning arms race, wherein chatbot developers vie with ostensible chatbot ‘detectors’ for technological supremacy. Such a competition will almost surely not be won by the good guys.”
Students are bragging that it’s easy for them to get away with this kind of cheating. Furthermore, professors will be leery of making a false accusation of cheating.
I think Hillard hits upon the truth in saying, “Thus, the alternatives before colleges are to enter a costly, never-ending, and largely futile arms race or to radically rethink how student learning is assessed. The former may be more likely given the higher-ed sector’s resistance to change. The latter is wiser.”