How ChatGPT Can Make Writing Easier — And What To Avoid – Forbes

Close up stock photograph of a mature man studying a see-through computer monitor that’s displaying … [+] text provided by an AI (artificial intelligence) chatbot.
The launch of ChatGPT 17 months ago is considered by many to be a technological inflection point as significant as the launch of the integrated circuit, the internet, and the iPhone. As with any such disruption, each novelty brings out legions of detractors and supporters.
To address both sides of the issue, let’s examine the positive aspects of ChatGPT and, as a cautionary tale, allay the concerns of the detractors.
As soon as ChatGPT came out, students started using the technology to do their written assignments which led to charges of cheating as this article by Michael Nietzel indicates. Businesspeople, too, started using the technology to do their daily work—emails, reports, marketing communications, product/service descriptions—and while businesspeople are not likely to be charged with cheating, the practice does have the downside of coming across as either bland, generic, or flowery. The title and subtitle of an article in The Atlantic says it all: “ChatGPT is Dumber Than You Think: Treat it like a toy not a tool.”
Instead, set an honor code for yourself. Never use generative AI tools to generate content from scratch. First, do a rough draft of your own. Then turn to the bot as a productivity tool and prompt it for a draft. Use any ideas or phrases from the bot and work them into a second draft of your own. Be sure you keep both drafts separate to maintain the integrity of your honor code.
Read that second draft, now focusing on the language. To heighten the accuracy and color of your description, query the bot again for synonymous words or idiomatic phrases. You can also use,, or Google’s Power Thesaurus, but if you give ChatGPT a sense of what you’re looking for, its natural language can provide closer suggestions. Don’t just cherry-pick from the list; use your native intuition to make the choice that fits best.
As you develop your text, look for additional information that supports and adds depth to your ideas. If the information you find is lengthy or complex, copy and paste it into the bot and prompt it to simplify. Read the simplification and then work the information into your original draft using your own words.
During the process, be sure to double check the accuracy of your research. ChatGPT is often not only inaccurate but as this New York Times article reports, can provide misinformation.
Continue to review your text. Remember that one of the most frequently cited adages among professional writers is “Writing is rewriting.” With each review, you should be implementing a practice known as “Spaced Learning,” or walking away from your text. It’s a human technique that predates ChatGPT. For context, the opposite of Spaced Learning is cramming—need I say more?
In using ChatGPT and other generative AI tools this way, you will be implementing “error-driven learning,” a practice drawn from technology that enables and improves human thinking. In a recent Wall Street Journal article, Professor Charan Ranganath who runs the Dynamic Memory Lab at the Center for Neuroscience at the University of California, Davis, references the work of two cognitive psychologist in 1990 who noted that “neural network models of artificial intelligence learn through trial and error.” Professor Ranganath then ran an experiment in his lab that “used neural network models to simulate what happens in the hippocampus—a critical brain area for rapid learning.” His results found “that the human brain can learn and retain far more through trial and error.”
So go ahead and use ChatGPT to help you generate, edit, enrich, and polish your content, but be sure to keep your hand on the tiller by using an equal measure of your own human capabilities.


Leave a Reply

The Future Is A.I. !
To top