Creativity Lite™
Field Notes
True birding field notes take time, patience, and creativity, combining sketches, notes about time of day, calls, behavior, environment, and so on. There’s beauty in learning to capture details about things we see in Nature. Sometimes - oftentimes - a painting or sketch of a bird can capture things about it that a photo may not. It also reinforces a deeper knowing and understanding of a thing and moves us towards mastery. It’s seen in the difference between paintings and sculptures of the human form done by those who understand human bone structure and musculature and those who don’t. There was a reason that da Vinci and Michelangelo spent time dissecting corpses.
Birding has come a long way from the days of weathered field guides, listening to bird songs on vinyl LP’s, wonky optics in binoculars, pencilled-in checklists, and field notes for the more dedicated. The addition of apps like eBird and Merlin give us tremendous resources for identification and have nascent gamification features that make birding a bit like collecting pokémon. We can already get plant identification natively in iOS: my guess would be that very soon we will have phone- and AI-enabled binoculars that can identify birds in real time (on the fly, as it were).
AI and Creativity Lite™
The spate of new generative AI tools reinforce this gap between the Lite™ version of creativity and a path towards mastery. They can either act as a tool to amplify our understanding and creativity or they can take us further away, acting as a crutch that makes us feel like we’re being creative without actually taking us on a creative journey.
In a sense all creativity is derivative - that’s not news to anybody. If I put together some chords and drum beats that I like it’s most likely close to a musical composition I’ve heard before. But I can still add my unique voice and intonation and instrumentation to it, make it my own addition to the world. Generative AI is much the same, except that its answers are derived from a much wider set of experiences. So wide in fact that we can ask it to “Write me a hit pop song” and it will do just that. This is where is becomes a crutch.
For context I present to you “Pudding Anarchy” (which is decidedly not a hit pop song). I can claim it as mine, but it was 100% written by Udio’s AI Music Generator. My only input was to prompt it for “a punk song about Banana Pudding”. As a screed against Big Pudding it really captures the sentiment of 1970’s antiestablishment punk.
Luddites, Die Hard 25, and Profits
Recently we saw studios proposing actor contracts that could use generative AI to replace much of the human aspect of what actors do. That same replacement is already taking place in non-union areas of the creative industry such as copywriting. In most ways that are important this is the 21st century’s version of Ned Ludd and his followers’ concerns “that cost-saving machinery … [would] replace skilled labor and drive down wages by producing inferior goods". The Luddites concerns about new machines in the textile industry were just part of a long-running battle that industry profits always wins.
It’s great that we might avoid seeing a virtual Bruce Willis in Die Hard 25, but ultimately the concessions gained by the actors related to the use of their digital likenesses in movies and AI training data won’t matter. The technology is already there for the studios to just create 100% virtual actors and skirt the whole issue. The question lies more in the ‘producing inferior goods’ part of the Luddites’ fears: will AI be used as a tool to amplify creativity, or will it just let a studio exec think they are creative.
The Creative’s choice
“But the only path to AI profitability lies … automation that turns the human in the loop into the crumple-zone for a robot”
— Cory Doctorow, Humans are not perfectly vigilant
I’m a fan of Cory Doctorow’s work and generally aspire to become a smart word-writer guy like him when I grow up. Perhaps with another 30 years of writing. Doctorow has taken generative AI out to the woodshed in a number of recent articles, especially as it relates to AI-related layoffs and the plight of the worker.
There are many, many problematic areas with this new generation of AI but, like those cotton-spinning machines of Ned Ludd’s time, it’s a tool that has too many benefits to ignore. I argue that as Creatives we should use it: for research, to quickly prototype ideas, to enhance our existing creativity … but we can’t just be lazy and stop there. The nature of generative AI is that it calculates the next most probable token (think “word”) based on the training data it was given. For the most part that training data is “everything on the internet”, so when we use generative AI to create songs, images, or text, what we get is an amalgamation of others’ work. This is anathema to having our own voice or putting our soul into our work.
Like using our senses and practice to identify birds, like those time-consuming field notes, our vision should be to go deeper towards mastery. It’s our job to not stop at the retelling and reskinning of the same old stories (or actors), having an AI generate complete songs, or watering down field observations into an online game. Just relying on them to make us appear more creative fails to show our soul to the rest of the world.