So your child came home and showed you a brilliant paper that he wrote on the Byzantine Empire. It was replete with stories of Emperor Justinian and his wife Theodora and their love of gold-laden garments, architecture, and the arts.
From behind his back, he pulled a small painting he had done in an impressionist style, using bright paint with layered textures highlighted by dabs of brilliant pastel colors. His subject? The beauty of Theodora. You paused briefly and wondered at his artistic accomplishment – but wondered why the Byzantine queen was holding a smartphone.
For a fleeting moment, you thought that Johnny, who until recently only was interested in soccer, wearing grubby white T-Shirts, cargo shorts, and playing Fortnite — now had an instant connection with the arts and embraced them with fervor.
But that evening you watched a news story about Openai, ChatGpt, and Dall-e2 and thought perhaps Johnny had not suddenly become a young man of culture – but rather a miscreant using the evil powers of technology to plagiarize his way into your heart and his teacher’s grade book.
Note: Each graphic in this article was created by using DALL-E2, the artificial intelligence engine developed by Openai.
Although the above scenario is a bit of hyperbole, such scenes might play out many times over the next few years as teens and tweens find their way through a myriad of new AI options. To some teens and tweens, this could conceivably ease the pain of academic writing or expressing their artistic side — even if they’re bereft of any discernable talent.
I recall having a similar conversation in the early 1980’s while I was working at a large computer company. I was hired as a writer and a director of video and film-based training programs for our customers and our staff of engineers. Word Processors such as WordPerfect, MS Works, MS Word, and WordStar had found their way onto the keypads and huge 1980’s CRT screens of our writing group.
In one case I recall vigorously debating whether WordStar’s spell and grammar checker was actually cheating. Although some of us probably felt that our jobs were in jeopardy – no one got replaced by a keyboard and a CRT screen. In fact, it made us more efficient.
The same was true with the development of products such as Photoshop, Illustrator, Corel DRAW, and many other graphics programs. After 40 years of such software on the market, they only made artists more efficient – and allowed them to bring the most challenging concepts to our eyes.
This is not to say that programs such as ChatGPT, including Meena, XLNet, GPT-2 and others will not cause issues in schools. The temptation to ask AI platforms to write a term paper is very real. Moreover, in some cases, the AI program will do an exceptional job. Initially, teachers and employers will find detecting such prose written by AI to be a challenge.
However, like software that can detect plagiarism — programs such as GPTZero are entering the market to detect writing created by AI. However, it is in its earliest stages and not nearly as accurate as programs that detect plagiarism. But their time will come.
AI programs may be good for our society. Educators might learn how to integrate AI into their classrooms – expanding how children learn.
Search engines will become much more than millions of links to articles related to your search. Instead, they will synthesize what can be learned by absorbing every available article related to your search and providing a concise, meaningful overview of the subject in question. That assumes that the data from which AI learns is truly accurate and balanced.
That said, there is no question, that today’s AI has built-in biases. I stress, that until that playing field is balanced, there will continue to be AI responses that tilt the information to one side or another — creating problems for society. It could represent FAKE NEWS 2.0.
I say this as someone that has tested ChatGPT, OPenAI and others for quite some time. In fact, as of this writing, we are just beginning to understand the weakness in the AI armor. This May 1st, 2023 article from Forbes provides a quick overview of the issue.
So, until there is better management of these new tools, take a deep breath. The world is not yet ending. Johnny has simply experimented with a powerful, but flawed new way of learning, albeit he’ll need some guide rails as to what constitutes his own writing.
I couldn’t think of a good way to end this article, so I did what any self-respecting man of the year 2023 would do, I asked ChatGPT to write the conclusion.
Here is the request that I gave to ChatGPT:
With humor, write a concluding paragraph on why parents should not be concerned that their child used chatgpt to write a report on the Byzantine Empire. Provide 1990 references that parents will relate to since most of them were in school at that time period.
Below is ChatGPT’s response:
“So parents, there’s no need to worry about your child using ChatGPT to write their report on the Byzantine Empire. After all, it’s like the ‘Google’ of the ancient world. Plus, ChatGPT is like having a virtual encyclopedia, like the World Book, but way cooler. And let’s be real, who wants to dig through a set of Britannica’s when you can have all the information you need at the tip of your fingers, like having an AOL search engine. Trust me, your child is in good hands with ChatGPT and they’ll be sure to get an A+ just like you did back in the 90’s.”
While it might not be the writing or wit of Mark Twain, with a little human creativity and editing, I’ll make due.
Note: If you are interested in A Wired Family presenting at your school, business or organization, simply email your request to: Marybeth@awiredfamily.org
