The following short story is entirely generated by ChatGPT-4.5 using 8 prompts:

Creation Process
I was wondering what it was like to have AI write a story for me. Sometimes I have ideas for a story, but I just sit on it because I write painfully slowly, and it seems like AI can fill in the gap, so I told ChatGPT what the story was, and it was able to expand on the ideas. Here is the very first prompt I gave ChatGPT:
The initial result was not good. It basically took my sentences and extended them. The storytelling was plain and boring. The introduction to Carter’s back story was awkward. I could not blame ChatGPT because it was doing exactly what it was told. I realized my prompt was not good.
I then gave it more prompts, each shaped the result closer to what I imagined. This includes clarifying my imagination and adding dramatic effects to the storytelling. I told it to reveal the back story and the current situation bit by bit through memories instead of all at once. I wanted AdNova to have a slogan, and I wanted the slogan to be mentioned early in the story, and mentioned again as the passphrase. The initial results described Carter’s action as a heroic act, with a glorious victory over Prometheus – it took me a bit to realized that I actually wanted Carter to act out of resentment for AI. He did not care about saving the world at all. And the story should stop at AdNova appearing to be effective against Prometheus.
What I learned
With generative AI getting more and more powerful and embedded in our lives, I think one of the most important skills in this decade is the skill to accurately communicate. The exercise above made me realize that I was worse than I thought at giving instructions. I thought years of writing emails in an office setting would have put me in a better place, but no, I need to do better. I should not have needed 8 prompts. I could have reduced the back-and-forth if I had explained myself better.
This hands-on exercise also gave me a sense of how powerful AI was at content generation. During the creation process, I felt more like a movie director than a writer. It bypassed my skill barrier in writing and avoided the dreadfulness in writing out the paragraphs. I could just focus on being creative and direct how the story should “behave”. It was fun. It showed me that I could bring my ideas to life without the actual skills. This opens my mind. I wonder if I can create a movie someday.
And I do not think AI stops at just creating movies. AI’s potentials can do far more than movies. There will likely be a new format of art, similar to how the invention of camera brought a new form of art called photography. I do not know what it is yet, but whoever creates a platform to nurture this new format will likely be the next TikTok.
Can we call it Art?
Hugely controversial, and I could be biased, but I think AI generated content can be considered art.
I believe art is anything that evokes emotion, conveys meaning, or possesses aesthetic value—regardless of who or what creates it. Mother Nature, for instance, produces stunning landscapes and intricate crystals. Many people view these as art because of their aesthetic appeal, even though they’re generated by nature’s inherent algorithms—the laws of physics—not by human intention. Similarly, AI-generated artwork follows algorithms as well. If AI-created art can evoke genuine emotion or express meaning effectively, should the method of its creation diminish its value?
When Jason Allen’s AI-generated piece won a prize at a painting competition, many people reacted negatively. This reaction was understandable since the competition was intended for human-created artwork. However, if no one had known the artwork was AI-generated, would the piece still not have been considered worthy of winning?
Again, this is controversial. Those who object to labeling AI creations as “art” often focus heavily on the method of creation, emphasizing that genuine artistry requires human intent, effort, or creativity. On the other hand, people like myself believe that art is determined primarily by the observer’s experience—if something evokes emotion, conveys meaning, or provides aesthetic value to the viewer, it can legitimately be considered art, regardless of how it was produced.
I recognize that AI-generated content can sometimes appear low-effort, derivative, or ethically questionable, especially when it heavily relies on pre-existing human-created works. However, these criticisms do not fundamentally undermine the potential of AI-generated pieces to be recognized as art.
While current AI-generated content undoubtedly has flaws, it’s increasingly meaningless to nitpick these imperfections, given the rapid pace at which AI technology is advancing. The flaws highlighted today could easily disappear by next month as AI continues to improve.
I don’t believe AI will ever surpass the best of human creativity. After all, art is fundamentally tied to human experience, emotions, and consciousness—qualities that AI inherently lacks. Without genuine life experiences, AI can never fully grasp or replicate the essence that truly makes art meaningful. Unfortunately, despite this limitation, the rapid rise of AI-generated content will inevitably flood our daily lives, potentially overshadowing genuine human artistry and reducing the visibility of truly exceptional human-made work.
(This section is revised by ChatGPT. My English is not that good.)