OpenAI has quietly released text-davinci-003, a new entry in the AI-powered GPT-3 family of language models, claiming that it can handle more complex prompts to produce longer results.
However, as Ars Technica (opens in a new tab) reports, enterprising users using Playground (opens in a new tab), the free GPT-3 offering, quickly discovered that the new model is even better for producing poetry and lyrics.
Hacker News commentators discovered (opens in a new tab) that he could write poems about Einstein's theory of relativity and then rewrite them in the style of the romantic poet John Keats. While Professor Ethan Mollick, a professor at the Wharton School of Business at the University of Pennsylvania, spoke lyrically about the possibilities in a series of tweets (opens in a new tab).
The GPT-3 generation of art and the future
The improvements in understanding of GPT-3's rhyme and meter are probably due to its being completed with even more reference material. The Github repository for GPT-3 (opens in a new tab) acknowledges that it derives its massive corpus from thousands of data sets.
Enthusiasts have noted (opens in a new tab) that previous iterations of GPT-3 had had a superficial understanding of rhyme schemes, but this latest step forward is a sign that this AI writer now has a complex understanding of meter and could possibly compose artistically. his own works.
While all of these developments are exciting, they raise the question of how human artists, writers, and, uh, journalists can exist alongside technology that is becoming less and less "cutting-edge" with each passing day.
Some fear that AI, which can type and retype faster than any human, will take paid work away from them. AI-powered text generation and manipulation is nothing new, with tools like Language is a Virus (opens in a new tab) and GPT-3-powered InferKit (opens in a new tab) offering a version of these functions for some time.
While it's true that AI text generators (and art generators like DALL·E) take a lot of the creative work, humans still have to create the prompts. And when it comes to the ability to regenerate parts of a human-dictated output (“repaint (opens in a new tab)”, which both DALL E and the alternative Stable Diffusion are capable of), c is also a human. - process driven.
So instead of seeing the latest generation of truly competent AI generation systems as a threat to human creativity, we might think about how it can adapt and collaborate with it.
It could be as a means of inspiration, making artistic processes accessible to more people, or creating new hybrid human/AI processes.
And if you still feel threatened, consider this: if the content you produce is playable by an AI system…wouldn't you rather create something else? A new and improved GPT-3 (or GPT-4, which is on the horizon) might give you more time to do this.