The Problem With AI Writing

While ChatGPT and its counterparts are revolutionary, AI writing will always long for the human touch.

Estimated reading time: 7 minutes

Unless you’ve isolated yourself in a snow-bound cabin to finish your book this winter, you’re aware of the explosive growth in Artificial Intelligence (IA) programs and applications.

Much of the focus has been on ChatGPT, which debuted in November. You can ask GPT to do pretty much anything. It can summarize long texts, create code, build websites, and — of potential concern to authors — create content. Lots and lots of content.

Up to now, we’ve had speculative futuristic AI scenarios ranging from the helpful Jarvis in the Iron Man franchise to the doomsday Skynet directing Terminators to wipe out the human race. For better or worse, the future is here. But what does this technological breakthrough mean for authors?

I’ve been getting sky-is-falling messages from writers who fear the new robot overlords will be putting words in blog posts, articles, and books. Breathe easy: for reasons listed below, I am convinced that ChatGPT and all its successors will never be able to replace writers. However, it doesn’t mean you should ignore the power of some of these sophisticated tools.

As revolutionary as the Internet

Tech titan Bill Gates believes ChatGPT and AI are as revolutionary as the microprocessor, the Internet, and the mobile phone. Gates’ old company Microsoft has poured billions into the R&D of AI, and they’re hardly alone. New AI tools are being introduced (at a rate of 200 per week) that can complete a mindboggling range of tasks, from creating personal travel itineraries (TripPlanner) to customizable royalty-free music (Beatoven).

But not everyone is comfortable with the warp-speed development of this technology. “People say books are the arteries of democracy; ideas spring from books,” says Mary Rasenberger, CEO of the Authors Guild, the nation’s oldest and largest professional organization for published writers.

In a recent article in Mashable, Rasenberger stated: “Unless [ChatGPT] is constrained, I hate to say this, but it will have a negative impact on the writing profession without a doubt and on the quality of what gets published. We need to protect book culture, but how do you do that in a world that is flooded with AI books?”

Runaway train

In late March, over 1,000 scientists and entrepreneurs, including Tesla CEO Elon Musk and Apple co-founder Steve Wozniak, wrote an open letter asking developers for an immediate moratorium on major AI experiments. “AI research and development should be refocused on making today’s powerful, state-of-the-art systems more accurate, safe, interpretable, transparent, robust, aligned, trustworthy, and loyal,” reads the letter.

All I can say is: Good luck. This runaway train has left the station.

The problem with AI writing

There’s no denying that ChatGPT4 and other content tools springing up daily are powerful. They can write poetry and essays. They can create entire websites based on a single paragraph and illustration. They’ve even passed several state bar exams. But I wouldn’t waste time dreaming of a world without lawyers — or nearly any other professional. There are serious flaws in these content-creation machines.

Is that a fact?

ChatGPT’s accuracy has been questioned in dozens of articles I’ve read. So I decided to put it to the test to help me write my author bio. The results were interesting, to say the least.

Steven Spatz is the CEO of BookBaby, a company that provides self-publishing services…

Well, that will come as a surprise to the real boss, CEO Tony van Veen. Plus, I’m retired.

Prior to joining BookBaby, Steven worked as Director of Sales at Free Spirit Publishing…

Who?

He also served as VP of Marketing at West Group Publishing and Exec. VP at Llewellyn Worldwide.

Wrong and wrong again.

Steven is an avid reader and writer. He holds a Bachelor of Arts degree in English from the University of Minnesota…

Woah. This has gone too far! I’m a proud graduate of THE University of Oregon! Thanks, but I’ll write my own bio.

Even unusual questions can reveal the limitations of the program. For example, if I asked my son, “Is England the capital of London,” he’d say, “No, it’s the other way around.”

Typing this question into ChatGPT4, I received:

No, London is not the capital of England. London is a city in England and the capital of England is London.

What do you know?

Developers explain that ChatGPT doesn’t really “know” things — it just produces text based on the patterns it was trained on. Because of the strangeness of the question, it doesn’t quite grasp what it’s being asked and ends up contradicting itself.

The takeaway? ChatGPT, in its current form, is loose with the truth! It’s intelligent in some ways but dumb in others. Remember, these AI tools can get stuff wrong — they even make things up. So, fact-checking remains a requirement for writing.

Who owns these words?

Using AI writing brings up two legal issues: plagiarism and copyright.

ChatGPT, in its current form, creates no genuinely original content. Instead, it builds on content that has already been written. The AI writing robot creates “new” content from different, combined text clusters. As a result, the finished text can be considered original in a legal sense.

Plagiarism occurs when a writer deliberately uses someone else’s intellectual property without consent or acknowledgment of the source. ChatGPT has no intellectual property as an artificial intelligence, which means you can use the texts without legal jeopardy.

Copyrights

Copyright law is another matter. For something to be protected by copyright, it must be an original creation of the human mind. This means it must result from human creativity, skill, or effort. The responses generated by ChatGPT are not considered original creations of the human mind and are not protected by copyright law.

You can expect many lawyers to focus on this area of case law. However, ChatGPT doesn’t list citations or attributions to original sources. “It is still unclear what the legal precedent may be for reuse of content if it was derived from the intellectual property of others,” said Gartner analyst Bern Elliot.

The human touch

ChatGPT and all its AI cousins are cold, heartless machines. They’ve never loved nor lost. Never experienced the thrill of victory or the agony of defeat. Unlike human writers, AI has not lived a human life and has no idea how the world works.

“It has no emotion, has no ability to think, no ability to generate any new ideas,” says The Author Guild’s Rasenberger, “We don’t want our arts to be limited to that. We want our arts to be human artists reflecting their current experiences and hopes and fears about our current world. That’s how arts have grounded civilization since the beginning of human civilization.”

Human writers add unique perspectives, facts, and creativity that AI writing lacks. In addition, writers inject personality, honesty, and — most importantly — emotions into their copy. I doubt whether ChatGPT or any AI program will ever learn to impersonate human qualities.

ChatGPT is simply regurgitating content that has already been written. Imagine if writers stopped writing. ChatGPT would be stuck reworking old words. There will always be a need for writers to create original content to feed this beast.

Harnessing the power of ChatGPT

Being of a certain age, I lived through the invention of the personal computer. At the time, many people worried that their jobs would be replaced by this fantastic new tool. Instead, computers revolutionized our society and created new opportunities. That’s how writers should regard the waves of new AI products coming into the marketplace.

Generative AI has the immense power to transform how we create. ChatGPT and AI can change how we brainstorm, write, edit, and publish. In my next BookBaby Blog post, I’ll identify the best ways writers can use these new AI tools for their own creative process.

source

(Visited 2 times, 1 visits today)

Z a c Zack