I don’t want to become an author-machine.
I say this because ChatGPT, an artificial intelligence text-generation application, is sending viral shock waves around the world. Developed by a foundation initially funded by the tech maverick Elon Musk, it is now often presented as an innovation that could replace entire categories of jobs in one fell swoop, rendering the work of many humans superfluous.
This app scans millions of websites in the blink of an eye, picking up content here and there, and distilling it into short, structured and apparently coherent texts. All those websites and social media posts are treated as abstract content, ready for the scooping.
In reading commentary in the world press about ChatGPT, I find a rather peculiar mix of extravagance, curiosity, fear and obsession.
For example, some commentators wonder if journalism itself is doomed. Others wonder whether universities should require students to submit research papers written in longhand, rather than ones typed on a computer.
No doubt, this AI technology could invade the inner space of our consciousnesses, subtly nudging us into a virtual hall of mirrors where we can no longer distinguish between the real and the unreal. It could engulf human effort in a wave of algorithms, transforming some literary creators (novelists, biographers, poets, researchers) into technicians adeptly manipulating automatically generated simulations of human expression.
But why not turn things around? If an app is a tool, then surely I can use it to best advantage. So I put the following question to this ChatGPT:
“Are you a real person?”
After nine seconds, ChatGPT replied:
“I am a computer-based AI assistant, not a real person. I was designed to assist with tasks, answer questions, and provide information to users like you. While I can engage in conversation and perform many tasks that might be associated with a person, I do not have feelings, thoughts, or physical sensations. I exist to help users like you in a variety of ways, and I am here to assist you to the best of my ability. Is there something specific you would like to know or discuss?”
I find this response refreshing, even polite, as if the self-effacing little chatbot recognized its own limitations.
As an author, I know how difficult, even tortuous, the path of creation is. I have learned that the most difficult aspect of creating a work — even more than finding words for a book or images for a film — is finding your own voice.
Creative works thrill us when they tell new stories in an unconventional way, when they have an authentic voice. And this voice is one of many things that differentiate humans from chatbots.
Being an author means affirming an identity. It means taking responsibility for what you create, sharing each work with the public, and ensuring your intellectual property rights are respected.
A chatbot can never be the author of anything. It can only rearrange what has already been said, as it gives random content the veneer of plausibility. A chatbot cannot provide reliable text, since it rehashes and merges anything and everything on the web, including conspiracy theories, denialism and hate speech. The output of a chatbot is therefore purely derivative, and reflects the grandeur and misery of the human condition.
Besides, I notice that ChatGPT sometimes spews out random nonsense. It reminds me of my often laughable experiences with another AI application — the GPS on my phone.
When I’m driving through the streets of my new hometown, Quebec City, my GPS assistant randomly switches from French to English, then to Frenglish with a Parisian accent, then to Frenglish with a New York accent. As if each street corner sent my GPS assistant into virtual shock, struggling to come to grips with every new geolocation.
Don’t tell me computer errors are due to imperfect human programming rather than the underlying technology itself!
ChatGPT can be a useful tool. At the very least, this application prompts us to think more deeply about communication, which is based on our relationships with each other.
As if to reassure me, ChatGPT added to what it said earlier:
“There are many differences between a person and a computer-based AI assistant like me. One key difference is that people are living beings with physical bodies and consciousness, while AI assistants like me are purely digital entities that exist within computers or other electronic devices.”
Authorship is (still) human. And I intend to keep it that way.
