One Teen’s Perspective on Writing with AI

Writing takes time, care, and skill. For a short piece of work, the time it takes from conception to completion can span hours to weeks. But for ChatGPT, writing a short story or poem takes less than a minute. In addition to the fast turnover, ChatGPT doesn’t exist behind a paywall. So, is writing with AI expected to be the norm?

I’m Cameron Alleyne, Write the World’s AI Liaison. I communicate how AI will become a larger part of our creative expression presently and in the future. When I look into AI, I think of how creative youth like myself will be impacted as we transition into higher education and the professional world. As a high school student, the typical AI use that I see is a peer using ChatGPT to write a research paper. Ten years in the future, by the time I have graduated from college and entered the workforce, AI use in my day-to-day life may be something I’m advised to take part in by other professionals. It concerns me now, because I only expect it to become more prevalent later.

How AI Writes

As WtW recently shared in a blog post, generative AI like ChatGPT “learns” from algorithms how to respond to requests from users. It is also able to expand on initial responses due to unsupervised machine learning, a process that enables AI to problem-solve using the information it has stored. 

This programming allows ChatGPT to communicate in a human-like way, making it useful in a professional setting or purely for entertainment. In either of these contexts, there is the question of whether or not writers should feel threatened by AI use.

The Concerns of Writers

Writers’ concerns about being replaced by AI were a focal point at the 2023 Writers Guild of America (WGA) strike. Writers in Hollywood who struggled to make ends meet because of how streaming changed the industry were confronted with the issue of their future in the writer’s room. 

Their central concern was that AI, which easily meets deadlines and doesn’t require compensation, may be more attractive to companies than writers themselves. After a grueling five months with those concerns luring over Hollywood writers, the WGA and the Alliance of Motion Picture and Television Producers (AMPTP) reached a resolution that is generally viewed as a win for writers: In short, production companies can introduce AI to the writers' room, but there are protections. For instance, companies must disclose if the material they provide to writers is AI-generated. 

Additionally, writers cannot be forced to use AI source material, and if they choose to use it, they do not have their rights to their project revoked. Most significantly, AI software cannot be credited as a writer, so it cannot be responsible for writing original material or rewriting original material from a writer. 

With these protections in place, the problem of AI should be resolved, right? 

Not exactly. 

AI-generated work is not eligible for copyright, making that content non-exclusive to the company. Additionally, since AI “learns” from existing work, the originality (or lack thereof) of the work produced by AI comes into question, making it a liability. This danger was modeled by the 2023 lawsuit against OpenAI, the parent company of ChatGPT. The Author’s Guild and 17 authors sued the company for allegedly using their novels, which are eligible for copyright, to teach ChatGPT.

Appropriate AI Writing

Another concern regarding AI is ethical use inside and outside of the writing sphere. Using AI, people can develop “deepfakes”: fake images or audio that emulate a real person’s likeness. 

Explicit deepfake images open the door to a new form of sexual harassment, with victims ranging from powerful female celebrities to underaged girls. So, AI can be used by users to produce material that is inappropriate or harmful in nature. Generative AI can also perpetuate racial bias, especially related to speech. A study published by researchers in arXiv, Cornell University’s research archive, that text using AAVE or African American Vernacular English was flagged as being indicative of laziness or stupidity. When creating characters or writing, AI may revert to incorporating stereotypes. Writing things that are unacceptable may give a person pause, but AI can produce it much quicker and without discernment between appropriate and inappropriate output.

Use of AI in Literature and Hollywood 

Although marrying AI with traditional writing can be daunting, it has already started seeping into the mainstream. For example, online writing assistants like Grammarly use AI to correct users’ grammar and tweak writers’ styles to make their work more refined or digestible. 

Writing produced using AI assistance, like the essay “Ghosts” by Vauhini Vara, has been well-received by critics and general audiences alike. Since AI “learns” from existing material, it can inadvertently help writers pinpoint cliches that appear in their or others' work; Charlie Brooker, the creator of the hit series Black Mirror, found that the quality of an episode written by AI for his show was poor, but it did bring overused plot points to his attention, which encouraged him to be more experimental. 

My Perspective 

As a student who wishes to continue writing,  I have worried about what the professional world will look like if I have to compete with or work alongside AI. It has left me to question what role the arts will have in our world by the time I’m an adult since anyone has the power to create using AI. I’ve realized that AI's abilities don’t have to threaten writers; instead, they can complement or build upon existing work. AI lacks human experiences and emotions. Writers lack the expansive database and rapid pace that AI possesses. Together, with people being the leading force, AI doesn’t have to be frightening. Instead, it can fit a niche in the writing world as a tool and not a replacement.