AI, or artificial intelligence, is the talk of the town. Every day, there are new products announced or launched, which boast of being driven by AI. Existing "smart" products are being slowly re-branded to "AI" products. As with the "smart" device hype, the moniker is being tacked on to products where it doesn't even make sense. Witness AI thermal paste! After the misses of blockchain and NFTs, people are excited by the apparent resilience of the AI hype. FOMO is a powerful drug and many are attempting to ride the new hype wave to fame and riches.
On the other side, creatives are crying foul over the apparent automation of what many considered the primary demarcator of consciousness - creativity. The "AI" being criticised is specifically generative AI (or Gen AI). Be it text generation, or text-to-image, text-to-video or text-to-audio, there is heated debate on the various implications of having artistic media being generated by just typing in what you want. Creatives are raging that this monstrosity was trained on existing creative material, without constent, and is being used to replace those that actually created the original. Non-creatives (for lack of a better term) relished at being able to generate profits from creative ventures, without having to actually deal with tempremental and expensive creatives.
Amidst all this are a bunch of people left scratching their heads at how we ended up automating the wrong thing. We thought machines would do all our work, leaving us free to pursue creative ventures. Instead, we toil to put bread on the table while the machines get to paint, make movies, and write books.
Before we delve any deeper, I want to highlight that there is no actual creativity that has been automated, so far. Fundamentally, you can think of Gen AI as "remix" machines. They can understand patterns from existing art and mix them together to create something seemingly new. Since the current Gen AI have been trained on a massive trove of human-created media, the amount of patterns it can pull from, would make it seem like something new to the vast majority who may not have the breadth and depth of exposure to various art forms from across the world.
At its core, this is the challenge we are dealing with. We have algorithms that can process and learn patterns from an unprecedented volume of media. With some feedback based fine tuning, it would be possible to generate media that is impossible for a human to reproduce. The shortened generation time also allows new processes in refining the final output. Already, human responses to media is being analyzed and algorithms being adapted to help nudge the responses in any direction needed. Imagine being able to generate movies for specific demographics that perfectly engage their intended audience. Unlike a real movie, this would (eventually) be generated in an incredibly short amount of time (let's say seconds). This can be tested with a focus group and their feedback can be used to adjust it further in a few more seconds, which would then be re-tested with a different focus group. In a single day, you can refine it to be an assured crowd favourite. What used to cost millions of dollars, with very little certaininity of success, has suddenly turned into a money printing machine. Is it any surprise that billions of dollars are being thrown into Gen AI?
Let's be honest. Regular human art will eventually be unable to compete with generated art. Human intuition on what would make art appealing to the audience, will always be hit-or-miss compared to a creative machine that can actually detect the response of its audience and adjust the output dynamically. Be it music, photos, movies, or paintings, the ability to actually measure how people respond to it, and then adjusting it to improve the response, is a superpower than will put generated art far ahead of human art.
However, a crucial aspect is being missed in this rush to replace human artists with generative AI is, sometimes literally, the human touch. In the midst of the mad gold rush to automate creativity, people seem to be forgetting that the meta of the art is often as important as the art. Sure, there will be many who consume art without caring about how or by whom it was made. However, the story surrounding the art, is often as important as the art itself. Art is not just media to be consumed. It is also a celebration of the human spirit and what it can achieve. Photos can tell a story beyond the canvas we see. We have movies about how other movies were made, because, ultimately, the creative process itself is a thing of wonder that has fascinated humanity since probably the first caveman picked up a piece of charcoal.
Every piece of human art, embeds a piece of the human in it. While it may not matter to everyone, I believe it will matter to enough people, for AI art to never be able to fully replace human art.