This post may contain affiliate links, which means I/we may receive a commission, at no extra cost to you, if you make a purchase through a link. Please see my full disclosure for further information.
In recent decades, the sciences have trampled and nearly extinguished the humanities in public schools and public opinion. We are told not to be artists or writers or musicians because we will never make a living that way. We should choose something practical, like economics or business to consider ourselves successful. Generative AI is no exception to this; but not only is it trampling over the arts, it is also putting the artists, writers, and musicians out of work. My argument is that using generative AI is not only unquestionably unethical but is also giving you a far inferior product than human writers.
The ethicality of using AI is questionable for several reasons. To begin with, by using AI instead of hiring a writer, you are putting dedicated craftspeople out of work. Writers have practiced and perfected their craft for years and by it are able to make a living by providing quality documents and manuscripts to businesses and people who do not or cannot write themselves. By using a computer to write your material, you aren’t just sidelining skilled workers, you are also ending up with a highly inferior product.
This inferiority stems from several problems with the technology. First, AI can only regurgitate what it has been fed, and what it’s been fed has been, in large part, is illegally obtained material. So, in addition to putting writers out of work, you are also using their previous work without permission or compensation. Last year, after the launch of Chat GPT-3, several lawsuits were filed against its creators, Open AI and Microsoft, for copyright infringement. The New York Times, a world-renowned newspaper, took legal action against OpenAI and Microsoft for the unauthorized use of its stories to train ChatGPT to answer users’ questions (“The New York Times Sues ChatGPT Creator OpenAI, Microsoft, for Copyright Infringement” 2023). In addition, seventeen best-selling and award-winning authors, including George R. R. Martin, David Baldacci, Jodi Picoult, and John Grisham, have also sued OpenAI for using their copyrighted written works to train the large language model (Romer 2023, News 2023). The intellectual property rights and copyrights of the product of these writers’ hard work and skill has been ignored, and they are rightly seeking damages. A third lawsuit was filed by eight U.S. newspapers, including the Denver Post, New York Daily News, and Chicago Tribune. They accuse both OpenAI and Microsoft of “‘purloining millions’ of copyrighted news articles without permissions or payment” in training their generative AI bots (“Eight US Newspapers Sue ChatGPT-Maker OpenAI and Microsoft for Copyright Infringement” 2024).
The other disadvantage is less about ethics and more about the quality of “writing” these generative AI programs give in response to your prompts. Artificial Intelligence is not, in fact, intelligent. It is merely trained to repeat the information it has been given, and it doesn’t actually even do that very well. GenAI is known for what have come to be called “hallucinations,” which happen when it gives a seemingly logical answer that is, in fact, entirely made up. Part of the NYT lawsuit includes instances when Chat GPT miscited the paper, giving the paper credit for reporting things that it never actually did. The paper is claiming “commercial and competitive injury” in that its reputation is damaged every time this misinformation is provided to users. If artificial “intelligence” is presenting and disseminating false information in a world already saturated with misinformation and false facts, is it really a technological breakthrough and positive contribution to society?
Another issue of quality is the actual “writing” itself. GenAI’s compositions are lacking in three distinct areas: the way it “articulates” its points, its faulty grammar and punctuation, and its incapability of understanding the incredible nuance of human language. One hallmark of ChatGPT writing is that it rephrases one point repeatedly while never developing an argument, and the supporting evidence it provides is generally sparse. It also has trouble with grammar, punctuation, and usage. Using Microsoft Word is standard practice for most people these days, and many writers find that, despite having been invented over forty years ago, the spelling and grammar check features give incorrect suggestions much of the time. If technology that has been developing since the 1970s still has not mastered language use, AI will most likely have the same insurmountable problem (Donlon 2024). As an editor, I know that running a piece of writing through an AI grammar check will always add more errors than it fixes. Using words in what could be considered nonstandard ways is also a common trip-up for technology. Many times, it has flagged my subject and verb agreement as incorrect when it wasn’t, and I haven’t even mentioned the problems that commas pose. Comma usage is highly problematic and, at times, controversial; many, if not most, people struggle knowing where and where not to use commas. When it comes to commas, much like for pirates, the rules are always actually just suggestions, so the idea of AI using them correctly is laughable.
Essentially, all this is to say that you should probably reconsider your use of generative AI. Hire a writer – we have been honing our use of language for most our lives and we want to put our knowledge and skill to use for you. As it has been said, just because science can do something, doesn’t mean it should, and the humanities are there to tell us why. As always, you get what you pay for, and GenAI is free.
Further Reading:
The Dark Side of GenAI: Exploitation and Compromise of GenAI Systems and Capabilities by Ivan Del Valle
Generative Artificial Intelligence: What Everyone Needs to Know® by Jerry Kaplan
Artificial Intelligence by Melanie Mitchell
The Alignment Problem: Machine Learning and Human Values by Brian Christian
The Coming Wave: Technology, Power, and the Twenty-first Century's Greatest Dilemma by Mustafa Suleyman
The Copyright Conundrum: AI and the Future of Creativity by Richard Anthony Aragon
Citations
“The New York Times Sues ChatGPT Creator OpenAI, Microsoft, for Copyright Infringement.” 2023. NPR. December 27, 2023. https://www.npr.org/2023/12/27/1221821750/new-york-times-sues-chatgpt-openai-microsoft-for-copyright-infringement.
Romer, Keith. 2023. “The Alleged Theft at the Heart of ChatGPT : Planet Money.” NPR. November 10, 2023. https://www.npr.org/2023/11/10/1197954613/openai-chatgpt-author-lawsuit-preston-martin-franzen-picoult.
News, A. B. C. 2023. “Famous Authors’ Lawsuit against Chatgpt Developer Gets Underway.” ABC News. November 29, 2023. https://abcnews.go.com/US/famous-authors-lawsuit-chatgpt-maker-openai-begins-initial/story?id=105239215.
“Eight US Newspapers Sue ChatGPT-Maker OpenAI and Microsoft for Copyright Infringement.” 2024. AP News. April 30, 2024. https://apnews.com/article/chatgpt-newspaper-copyright-lawsuit-openai-microsoft-2d5f52d1a720e0a8fa6910dfd59584a9.
Donlon, Marie. 2024. Review of A Basic Understanding of Spell-Checking and Grammar-Checking Algorithms. Insights.globalspec.com. March 20, 2024. https://insights.globalspec.com/article/11290/a-basic-understanding-of-spell-checking-and-grammar-checking-algorithms#:~:text=In%201971%2C%20a%20graduate%20student%20under%20Earnest%2C%20Ralph,offering%20no%20suggestions%20for%20fixing%20the%20flagged%20words
Comments