Thanks for reading Actual Intelligence with Dr. Steve Pearlman! Subscribe FREE to receive new posts and support my work.

APA to Students: Don't Bother to Think for Yourselves Anymore. Let AI Do It.

If in the future you want a psychologist who can actually think about psychology, or a doctor who can actually think about medicine, or a teacher who can think about what their teaching, or a lawyer who can actually think about the law, then the new American Psychological Association’s (APA) A.I. policies should make you concerned. Maybe they should even make you angry.

As many who’ve been to college already know, the APA’s standards for what constitutes academic integrity and citing sources is the prevailing standard at most institutions. When students write papers or conduct any research, it’s typically the APA’s standards that they observe for what they are permitted to use and how they must disclose their use of it.

Yet, when it comes to supporting critical thinking and actual intelligence, the APA’s new standards just took a problematic if not catastrophic turn. And the irony is palpable. Of all the organizations that set standards for how students should use their brains, you might think that the American Psychological Association would want to hold the line in favor of actual thinking skills. You might think that with all of the emerging research on A.I.’s negative consequences for the brain—including the recent MIT study that showed arrested brain development for students using A.I. to write, which you can learn more about on my recent podcast—that the APA would adopt a vanguard position against replacing critical thinking with A.I. You might think that the APA would want to bolster actual intelligence, independent thought, evidence-based reasoning, etc. But instead of supporting those integral aspects of healthy brain development, the APA just took a big step in the opposite direction.

I’m referring to the APA’s new so-called “standards” for “Generative A.I. Use,” standards that open the doors for students to let Generative A.I. do their thinking for them. For example, the APA liscenses students to have A.I. “analyze, refine, format, or visualize data” instead of doing it themselves, provided, of course, that they just disclose “the tool used and the number of iterations” of outputs. Similarly, the APA welcomes students to have A.I. “write or draft manuscript content” for them, provided that they disclose the “prompts and tools used.”

To be clear, the APA’s new standards make it all too clear that it is very concerned that students properly attribute their uses of Generative A.I., but the American Psychological Association is not concerned about students using Generative A.I. to do their thinking for them. In other words, the APA has effectually established that it is okay if students don’t analyze their own data, find their own sources, write their own papers, create research designs, or effectively do any thinking of their own; it’s just not okay if students don’t disclose it. In short, the leading and most common vanguard for the integrity of individual intellectual work just undermined the fundamental premise of education itself.

What the APA could have done and should have done instead was to take a Gibraltarian stand against students using A.I. in place of their own critical thinking and independent thought. That is what it has done to this point. For example, students were simply not permitted to have a friend draft an essay for them. They were not, in many circles, they were not permitted to allow a friend to proofread their work unless the syllabus licensed them to do so. But for some reason, since it is an A.I. drafting the paper instead of a friend, the APA considers it permissible.

Thanks for reading Actual Intelligence with Dr. Steve Pearlman! Subscribe free to receive new posts and support my work.

Consistent with its history of guarding academic standards, the APA could have said that students who have an A.I. “analyze … data” or “write or draft manuscript content” were not using their own intellect and therefore cheating. Period. Doing so would have sent a strong message across all of academia that permitting students to use Generative Artificial Intelligence instead of their actual intelligence was a violation of academic integrity, not to mention a gross violation of the most fundamental premise of education itself: the cultivation the student’s mind.

To be fair, not all of the usages of A.I. referenced by the APA’s new standards are cheating. For example, allowing students to use A.I. to “create … tables” or “figures” instead of painstakingly trying to build them in Microsoft word, would not replace the student’s meaningful cognitive work.

Furthermore, and more importantly, the APA’s policies are not binding. Educators, departments, and/or institutions need not follow suit. Any given educator can still restrict A.I. usages and determine their own standards for what is acceptable in a given course, including the establishment of policies that would treat using A.I. to “analyze … data” as cheating (which it should be).

And finally, the APA still asserts that “AI cannot be named as an author on an APA scholarly publication.” Yet, to co-opt a psychological term, that seems nothing if not “schizophrenic.” After all, if a student uses A.I. to find its resources, “analyze” their “data,” and “write” their “manuscript,” then why shouldn’t it be listed as an author, if not the lead author? What, after all, is the student really doing anyway?

Thus, as arguably the leading force for what constitutes academic integrity vs. cheating, the APA’s move at least implicitly licenses students across academia to use Generative A.I. in ways that will undermine their individual work, critical thinking, and overall actual intelligence. Once again, the APA just told students everywhere that using A.I. to “write or draft manuscript content” for them, instead of thinking about it themselves, developing their ideas themselves, referencing sources for themselves, perhaps even reading sources for themselves, and on and on, is perfectly okay as long as they cite it when they do so.

And while it remains true that faculty can do as they wish, imagine being that high school, college, or graduate school educator who has to stand against the APA. Imagine having to hold the line against what will be mounting droves of students who ask, “Why can’t we use A.I. in your class when we use it in our other classes?” And who ask, “Why can’t we use A.I. in your class when the American Psychological Association says it is fine?” Considering that educators with stricter A.I. policies are already seeing students unenroll from their courses, the new APA standards my prove catastrophic.

So, that returns us to the emerging problem: If you think that academic institutions should graduate students who can think critically about their subject of “expertise”—if you want a doctor who can think about medical things—then the APA just told you that you had better thing again.

(This article written with no Artificial Intelligence, only the actual kind.)

If you support actual intelligence, please share this with other likeminded people.

***

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit pearlmanactualintelligence.substack.com

382 - La méthode "Make Time" de John Zeratsky et Jake Knapp (en 5 idées simples)