Sometimes, fads sweep the nation, taking the country by storm before fading away into relative obscurity (or, at least, a new, smaller stability). They might be toys that children adore before they toss it aside for the next generational interest, like Webkinz, Hexbugs and Fidget Spinners. They may come in the form of some technological product, like Google Glass, Smell-O-Vision and (arguably) 3D movies.
Or sometimes, they cease to be fads at all, quickly incorporating into mainstream society—becoming a part of our daily lives.
We’ve yet to see what side of the line OpenAI’s Chat Generative Pre-training Transformer, more commonly known as ChatGPT, will fall.
In a way, you might describe ChatGPT as if Google was able to hold a conversation. The software is a deep learning model that weighs the various words of an input prompt in order to provide an answer that mimics how a human might respond. Put simply, the software doesn’t know anything—it’s just quite good at both using an input to find and synthesize the relevant information as well as providing it in a conversational way.
And because of that conversational feature, ChatGPT goes much further than any old query for “best poems by Edgar Allan Poe.” Instead, you can ask it to “write a poem in the style of Edgar Allan Poe,” and it would deliver. The software can summarize entire articles, write blog posts in favor of or against topics, create movie scripts and much more. It even offers a basic understanding of Christian theology.
To give you a better idea of that, here’s two screenshots. The first shows ChatGPT’s response when asked to provide a paragraph in support of watching movies, and the second shows its response when I asked for a paragraph opposing movie watching.
While the answers are a bit clunky, you’d be forgiven for not realizing that each had been generated by an AI had you not been informed of that fact prior to reading.
So, the $64 question is “What questions can you ask ChatGPT?” The $64,000 question is “How might ChatGPT change the way we work or learn?”
Well, first, let’s acknowledge that the software seems to be here to stay. At the very least, major tech companies seem to think so, as Microsoft, Alphabet (Google’s corporate umbrella) and Chinese tech behemoth Baidu have all made intentional changes to their respective search engines in response to ChatGPT’s public release in Nov. 2022, according to Wired.
It’s certainly on the radar in the academic field, too. According to Nature, the AI has already been listed as a co-author on “at least four…published papers and preprints.” A business professor at the University of Pennsylvania found that the software would pass a Master of Business Administration exam with a “B to B- grade on the exam.” And when The Stanford Daily, a student-run newspaper at Stanford University, conducted a poll regarding ChatGPT, they found that 17% of nearly 4,500 students had used the software to help in finishing “fall quarter assignments and exams.”
So yes, ChatGPT leaves some concerns when it comes to academia, and students shouldn’t be surprised if their school district blocks the software on school devices. For instance, some have voiced concerns that the complexity of the software has all but obliterated the take-home essay. Others point to the fact that the software readily admits that ChatGPT might produce inaccurate data. And, on the ethical side of things, we must contend with the morality of a student passing off generated (and thus plagiarized) work as his or her own.
But the tool, since it’s essentially a more intense form of search engine, can also be used to benefit someone’s education, too. For instance, the software can be used to help students come up with ideas for school projects, offer up counterarguments that students can argue against and help refine an awkward sentence that a student is having trouble workshopping. A single query in ChatGPT might procure helpful data for students to verify and use that would otherwise take countless hours to find. And as for that issue about students submitting AI-generated essays, OpenAI has released its own AI Text Classifier that analyzes input text to determine whether it was written by an AI or a human. (While not perfect, Plugged In’s own testing of the tool resulted in relatively accurate results.)
So how should parents approach AI software such as ChatGPT with their children? It’s a mixed bag, one that shouldn’t be wholeheartedly affirmed nor immediately dismissed.
Like any piece of technology, ChatGPT comes with legitimate concerns, and parents should emphasize the importance of putting in the work to truly learn a topic rather than using the software to find easy answers. But used properly, it also provides ways for students to bolster their studies without compromising their ability to learn.
In that respect, ChatGPT would seem to share the same benefits and drawbacks of many educational, technological tools. Calculators were once wildly controversial tools in the classroom, thought by some to short-circuit a student’s understanding of mathematical concepts. The internet itself threw open the doors to dizzying new information sources—and dizzying new ways to cheat. Tools, including technological tools, are rarely evils in themselves. It’s how we used them that matters.
Ultimately, it’ll be up to each family unit to determine if and how they’d like to use the software—for now—but we’d caution against burning the house down to smoke out the rat. If it’s a fad, it will inherently fade. But if it becomes part of our educational world and, by extension, part of our society, it’s important for moms and dads to guide their kids’ usage as best they can—teaching them the difference between wise use and wholesale abuse.