ChatGPT, Artificial Intelligence, and the Future of Writing | Psychology Today

The potential of Artificial Intelligence (AI) to revolutionize writing education is immense. AI can use natural language processing (NLP) to help students analyze their writing in ways that teachers have never been able to do before. AI systems can evaluate structure, grammar, tone and semantic meaning of content to provide personalized feedback to students. In addition, AI systems can quickly scan content for plagiarism and other errors, enabling students to self-correct in real-time.

AI can also be used as an effective tool for making writing more interactive and engaging for students. AI-enabled systems can make writing assignments more tailored to individual student needs and interests. For example, AI-enabled essay writing platforms can generate real-time prompts and guide students through the writing process. AI can even help generate writing assignments that are adapted to the individual’s skill level and interests. For example, AI can generate writing topics based on the student’s previous writing samples or past topics that the student has enjoyed writing about.

AI can also help with assessment and grading, shifting the focus from correcting errors to fostering better writing skills. By leveraging AI technologies such as machine learning, AI-enabled writing assessment tools can automatically evaluate and grade essays more accurately and efficiently than manual grading. AI systems can also provide students with detailed feedback on their writing, which can help them further improve their writing skills.

Overall, the potential of AI to transform writing education is immense.

Guess what? I didn’t write any of the above! That was fully written by a robot via the newly developed artificial intelligence software, ChatGPT. Using this software, I requested a 200-word essay on how AI technologies will affect writing education in the future.

The robot wrote that piece in about 20 seconds.

While the answer is reasonably well-written and on-point—and while this new technology is, well, actually pretty amazing at what it does, along with educators all over the world, I see some pretty glaring problems with this software.

Interestingly, when I asked the same software to write a 200-word essay regarding how AI relates to plagiarism, the result it gave me was total garbage—advertisements typed up by the robot, in fact. I tried this on three different occasions and the same outcome happened each time. I am not sure if this fact is somehow intentional or not. But it certainly is noteworthy. It comes across to me as if the developers of this software designed the software to not broach that particular subject. This said, as an educator who mentors students in writing extensively, I see this issue as more than a little problematic.

How chatGPT Generally Works

It seems that chatGPT has emerged on the world scene in the blink of an eye. So many folks reading this may not yet be aware of what this software does. Basically, by using extraordinary power in quickly searching content online, coupled with surprisingly good grammatical and writing skills, this software can write original content on pretty much anything. In pretty much any style.

You can ask it to write a 500-word essay, for instance, on the neurobiology of human emotion. Or the main themes of the works of Charles Dickens. Or the history of evolutionary scholarship. Or…well…just about anything.

Further, you can ask for any format of writing. Haikus, limericks, 500-page manuscripts, 8,000-word essays, etc. Further, you can ask it to produce content in a broad array of styles.

My fiancee, Shannon, who works in the field of content direction, showed me various eerie examples. She asked it to write a 500-word essay about cars. And she did this multiple times. On one occasion, she asked it to write this in the style of Snoop Dog. And wow, was it good. Then she switched to the style of Mark Twain. And it sounded just like his writing. Then she switched to the style of myself—asking for a 500-word essay about cars in the style of Glenn Geher. It used words that I often use in my writing (e.g., genuinely and simply). It used punctuation that I often use, such as em dashes and semicolons. I have to say, even I could have been fooled into thinking that I’d written it. This software is that good.

Potential Problems with AI-Assisted Writing Software

While ChatGPT does not seem extremely willing to discuss potential problems with this software, I, on the other hand, am more than happy to do so. In fact, as you’ll see, although I rarely consider myself an alarmist, I am reasonably convinced that the ability of humans to communicate with one another—especially in written formats—is going to be adversely affected. Permanently.

From an educator’s perspective, here are three potential problems associated with AI-assisted writing software:

My Anti-AI Policy

As a relatively proactive teacher, I actually worked to include anti-AI statements in my syllabi for two classes that I am teaching this winter (less than a week after finding out about ChatGPT). I fully realize that any policies that professors create on this issue—especially at this stage, where this stuff is all brand new—are equivalent to sticking your finger in the hole of a dam that is necessarily going to burst. But, as always, I try my best to deal with problems as they arise.

My approach to this issue, for now, is to rely on two important principles. First, I emphasize to students just how deeply important it is for them to develop writing skills in college. Second, I rely on their integrity, by asking them to indicate, in writing, that their samples of written materials are original.

If it’s helpful for anyone reading this, my specific policies on this issue are in the references below.

Note that I am not the only one developing policies surrounding this issue. In a highly publicized recent development, the New York City Public School District announced that it is blocking and disallowing the use of ChatGPT in its schools.

A Finger in the Dam

As I indicate above, I realize that policies such as the ones described above are, at best, equivalent to putting a finger in a dam that is about to burst. The temptation to utilize software that can immediately write reasonably good papers on nearly any topic—of any length and in any style—is simply too strong. And, unlike my usual rose-colored-glasses self, I am not (at this moment) too optimistic about the future of writing education and of writing in general.

Bottom Line

It’s a new world. The robots are here. As was predicted by all kinds of sages of the past (such as Kurt Vonnegut via his classic Player Piano), they are not going away and they are changing the human experience profoundly. In much of my past writing, I’ve explored how many facets of the modern world are mismatched from the ancestral conditions that our ancestors evolved to experience (see my and Nicole Wedberg’s book, Positive Evolutionary Psychology: Darwin’s Guide to Living a Richer Life). If having robots write term papers and theses is not mismatched from our ancestral lifestyles, I don’t know what is.

Generally, I am not a fan of pointing out problems without concurrently pointing out suggested solutions. In that spirit, I suggest that educators, students, government officials, and leaders across all industries think carefully about this issue, including its potential effects on the ability for people to write and to think. I am not sure if there is a solution out there. But hopefully working together on this issue, the global community, with lots of shared interests in the future, will figure it out.

Note: Thanks to Shannon Guyton for first pointing this software out to me and for helping me see many of the issues associated with this software. Thanks also to Marshall Siegel who had various helpful suggestions regarding this piece.