Much has been said about large language models (LLMs) in 2026. From its environmental implications to how it can benefit the globe, public discourse has become extremely mixed in opinion.
Especially for writers and artists, artificial intelligence carries a unique subset of concerns. In a constantly modernizing techno-state, we are struggling to ground ourselves in a stable creative climate. As writers, how can we navigate a world that seems to challenge our identities? What can we do when a new technology seems to threaten our craft?
- Philosophy of Writing
“Everybody is talented because everybody who is human has something to express.”
— Brenda Ueland (Mossflower)
The writing is a reflection of the writer. Regardless of domain––whether it be journalism, novel writing, poetry, etc.––our arrangement of words on the page communicates something about lived experience. Synthesizing a work can be both meditative and fulfilling. It is also extremely difficult.
Over the centuries, the discipline of writing has evolved tremendously due to technological advancements. Storytelling originated by means of oral tradition, long before paper was popularized. Today, it is an industrywide standard to finalize works through apps. AI is simply another change in the path we take for creation. However, it is wholly different in that the former technological innovations never questioned the necessity of writing by humans; up to this point, it has been an explicitly human capability. Now, technology has colonized even this realm of personhood.
In popular culture, we often think of writing as a product. We see it in its fully realized state, and we praise it and critique it that way. Rarely if ever do we peel back the layers and examine the process: one work is the result of hours of research, contemplation, and struggle. It is an amalgamation of epiphanies and obstacles. Yet the way we are celebrated is in terms of output; our works are read, related to, and discussed, but the process we took to get there has no cultural capital.
This is one predominant reason why LLMs have been able to infiltrate writing spaces. One article from New York University (NYU) reads, “AI can … generate unique concepts that might inspire an artist’s next masterpiece” (Fortino). Using AI to brainstorm seems to be widely accepted, even in many academic circles. Because we envision the value of writing as a final result, the process we take to get there holds less importance. Therefore, tech can break into the procedure without it delegitimizing the product.
The process of writing itself is an expression of humanity. As people, we all think deeply about something. The period of reflection where we draw outlines or ponder how we can communicate something simply holds significance in a way that instantaneous generation cannot. People who are content with an AI generated story misconstrue the purpose of interacting with writing as consumption. In reality, reading is a conversation with the audience and writer: an infinite series of arguments, discussions, and opinions. All writing is elastic. All writing shapeshifts. There is no value in an AI generating “thoughts” on a movie, or a critique of pop culture, or even a fictional story about human experience. The purpose of reading is to get a glimpse of someone else’s world, their story, and to learn how we can bring a piece of it back into our own lives. LLMs, no matter how good they are at replicating this human experience, will never hold that sentimental value.
There are some things about humanity that cannot be objectified with scientific language. The importance of writing is deeply personal, spiritual, and intrinsic, something no machine could ontologically achieve.
- Purchasing Our Freedom
“Can AI be trusted?” a voice asked as I watched from the theater seat. I was intrigued. Amongst the trailers of PG-13 adaptations, sequels, and biopics, this odd video began to play. Beneath a grainy filter and staccato piano, the footage was interspersed with moments of familial joy, nature, and scientific discovery. Disembodied voices blurted out complex moral questions. Only by the end, when the Claude logo emerged from its salmon background did I realize it was an advertisement for Anthropic’s LLM (“There’s Hope in Hard Questions”).
Anthropic’s ad speaks to a new phenomenon emerging in AI discourse. Recently, a group of mostly liberal and left-wing skepticists have carved out a place in conversations, warning about AI’s ramifications and threats to human functions. These voices tap into a popular sentiment—according to a 2025 survey, half of American adults say the use of AI in everyday life makes them concerned, and 53% say it will worsen creative thinking skills (Faverio and Kikuchi).
To pander to a reluctant audience, companies now offer something that is universally desirable—freedom. Specifically, the freedom to pursue your passions without inhibitions. Google promises that its planning tools will release you from the banality of officework so you can make time for what you love. Claude claims that instead of replacing the human, it will perform your robotic responsibilities to free up your schedule. It also boasts its increased ethicality compared to other models such as Grok or ChatGPT. But Anthropic doesn’t care about your passion project. It doesn’t care about community or making a better world for scientific discovery. By co-opting the aesthetics of artists, and moreover, of humanity—like the grainy cinematography and quirky score in the ad—they disguise a product as the organic solution.
These companies do get one thing right: our freedom is constrained by office jobs and online work. Our soul craves the pursuit of humanity and passion, not the mundanity of emails, essays, or finance reports. But the solution we want isn’t something we can acquire by using their AI. It’s an issue that is intrinsic in the architecture of our economy. It is the system which sets the value of artists lower than monotonous productivity. It is the system which forces us to commodify our art and assign a quantitative “price” to pieces. A tool which lets us generate fire breathing dragons or sets up a Wix site for our portfolio won’t give us the actual freedom we want. So long as our institutions continue to see humans as consumers and workers as tools, we won’t receive the liberation we crave.
- The Human
AI has forever changed how we think about writing. It has infiltrated the thing we find solace in, the tool we use to understand ourselves. To see people around us, including those we love, using LLMs feels like a betrayal. But what can we actually do about it?
If you are anything like me, you feel a great deal of anger. This tech is attacking something we cherish. Not only that, but the increasing popularity in AI models, especially in younger generations, is something that is devastatingly apparent. Many of our peers offload their tasks to LLMs, relying on it to produce study materials or form outlines for their essays.
My initial instinct was to shame these people; this is also the most common response in most anti-AI spaces. “AI shaming” is now so common, even in academia, that research has been done to examine the characteristics and results (Giray). It is an understandable impulse to humiliate or guilt people that enable this technology to thrive. After all, LLMs would not be this successful if not for their complacency. Our resentment comes from very real places. For example, reducing intellectual engagement and creative friction, which are effects of AI use, have neurological implications (Kos’myna). Anger, in this scenario, is warranted—but it is not the solution.
The reality we need to face as writers who oppose LLMs is that they are already in widespread usage. It is still powerful for us to boycott it, and not treat it as an inevitability. Yet, the fact that much of the online population is already using AI is something that requires solutions with tangible impact, not hostility.
In fact, research has proven that shaming is an ineffective tactic to address nearly any issue. Psychological reactance theory (PRT) is a widely supported phenomenon by researchers. It posits that when something or someone tries to stop a person from doing an action, it makes them want to do it more. They experience “psychological reactance”, believing a part of their freedom has been compromised. As a result, they will be motivated to act against the boundary, towards “freedom restoration” (Rosenberg and Siegel). Shaming or reprimanding those already with an AI habit, therefore, might actually exacerbate the issue instead of making them stop.
The alternative to shaming is harm reduction. Harm reduction is an approach first applied to adults with substance abuse. It was conceived with the understanding that full abstinence was an unrealistic goal for most patients. Since then, it has been widely adopted in schools as a way to decrease adolescent drug and alcohol use. Instead of relying on ineffective programs like D.A.R.E. (drug abuse resistance education, which told kids to “just say no” to drugs. It inadvertently raised the levels of adolescent use instead of mitigating it), education is now more focused on raising awareness on the downsides of substances, as well as how to consume them responsibly (Canadian Paediatric Society). I believe this same framework can be applied to AI. Instead of blatantly prohibiting its usage, students should be taught the ways in which AI can be both harmful and helpful. It might seem counterintuitive, but teaching people the right way to use AI might decrease the amount they actually end up using it.
Research has already been done to back this theory, and companies don’t want people to know about it. Six studies and additional data found that those “with lower AI literacy are typically more receptive to AI”. Contrary to expectations, this receptivity is not explained by people’s opinions on AI ethics, ability, or impact on humanity. Instead, it is because people with lower AI literacy are more likely to see it as magical and be awed by its abilities. The paper goes as far as to suggest that companies should market to consumers with lower literacy, as they are more susceptible to branding strategies (Appel et al.).
What this indicates to us is that AI literacy is more important than ever in fighting the crisis. When people are informed about what AI is bad at, they are less likely to use it for those things. Harm reduction, having conversations with AI users, and raising awareness are the most effective ways to combat AI overreliance. We can’t stop people from using LLMs as long as they’re publicly available—but we can try to educate them as much as we can.
- Art Is Forever
It’s a difficult time for many artists and writers, but I would like to think there is still hope.
While some Hollywood titans like Martin Scorsese and A24 have come to adopt the technology, many others are taking a firm stance against it. In 2024, Hollywood writers went on strike against generative AI, and won (Kinder). The victory means a lot for all creatives. It shows we can fight back.
Culture is incredibly significant in our lives. We connect with each other on our favorite movies, attend concerts together, join book clubs, and discuss fashion shows. AI can’t annex that part of life founded upon the shared enjoyment of human expression. It will never know the catharsis of sitting down and spilling out your thoughts on paper. It will never laugh at a restaurant with its friends or try its grandma’s soup or sift through memories of its childhood.
The human craving for art and connection will never disintegrate. AI is a lot of things, but it will never be a writer.
Works Cited
Appel, et al. “Lower Artificial Intelligence Literacy Predicts Greater AI Receptivity.” 2025.
Becker, Adam. “The Useful Idiots of AI Doomsaying.” 19 Sept. 2025.
Canadian Paediatric Society. “Harm Reduction: An Approach to Reducing Risky Health Behaviours in Adolescents.” Paediatrics & Child Health, vol. 13, no. 1, Jan. 2018, pp. 53–56, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2528824/, 10.1093/pch/13.1.53. Accessed 21 July 2026.
“Claude’s Constitution.” Anthropic.Com, Anthropic, 2025, https://www.anthropic.com/constitution. Accessed 20 July 2026.
Faverio, Michelle, and Kikuchi, Emma. “Key Findings About How Americans View Artificial Intelligence.” 12 Mar. 2026.
Fortino, Andres. “Embracing Creativity: How AI Can Enhance the Creative Process.” Sps.Nyu.Edu, 2024, https://www.sps.nyu.edu/about/news-and-ideas/articles/etc/2024/embracing-creativity-how-ai-can-enhance-the-creative-process.html. Accessed 13 July 2026.
Giray L. AI Shaming: The Silent Stigma among Academic Writers and Researchers. Ann Biomed Eng. 2024 Sep;52(9):2319-2324. doi: 10.1007/s10439-024-03582-1. Epub 2024 Jul 8. PMID: 38977530.
Habib, Sabrina. “AI Can Help − and Hurt − Student Creativity.” University of South Carolina, 5 Feb. 2024, https://sc.edu/uofsc/posts/2024/02/conversation-ai-help.php. Accessed 13 July 2026.
Kinder, Molly. “Hollywood Writers Went on Strike to Protect Their Livelihoods from Generative AI. Their Remarkable Victory Matters for All Workers..” The Brookings Institution, 12 Apr. 2024.
Kos’myna, Nataliya. “Your Brain on ChatGPT: Accumulation of Cognitive Debt When Using an AI Assistant for Essay Writing Task – MIT Media Lab.” MIT Fluid Interfaces, 10 June 2025.
Mossflower. “12 Quotes WE LOVE about Writing.” Mossflower.Com, Reading & Writing Project at Mossflower, 2 July 2024, https://mossflower.com/blog/12-quotes-we-love-about-writing/. Accessed 13 July 2026.
Rosenberg, Benjamin D., and Siegel, Jason T. “A 50-Year Review of Psychological Reactance Theory: Do Not Read This Article..” Motivation Science, vol. 4, no. 4, Dec. 2018, pp. 281–300, 10.1037/mot0000091.
“There’s Hope in Hard Questions.” 9 July 2026.
Thoreau, Henry David. Walden. London Vintage, 9 Aug. 1854.
