How AI will improve human art and writing
Don't panic. It's likely that AI is the best thing that ever happened to human-generated content.
Panic is in the air.
San Francisco-based Open AI made cutting-edge generative AI accessible to the public. The public tried it and realized that publishable posts, articles, essays and novels can be created by AI in seconds. AI, as an “artist,” can churn out award-winning “art” in vast quantities.
CNET Money has published dozens of news articles written substantially by ChatGPT, and nobody noticed until the fact was reported. Bloggers and marketers are actively using ChatGPT to generate content. This is only the beginning. Within months, AI-generated written content will be everywhere. And in just a couple of years, the vast majority of online content is likely to be synthetic media — content generated in part or in full by AI.
And that’s why people are panicking.
The existence of these tools represents the “collapse of the creative process” and even the “end of writing,” according to some.
“Teachers are worried” and have been warned to expect a “flood of cheating.” The New York Education Department blocked ChatGPT on all school networks.
The future looks grim for human artists and writers. Or does it?
We’ve been here before
In 1826, Joseph Nicéphore Niépce took the world’s first photograph. Within 23 years, some 100,000 Parisians were getting their picture taken at least annually. In the 1880s, George Eastman created film in rolls, enabling an eventual mass market in cameras, which turned everyone into a photographer.
Within a few generations, getting one’s picture taken became a banality in the industrialized world— something everyone did.
What really changed was the medium photography largely replaced for portraits — painting.
Before photography, most painting existed to capture people in painted portraits. It also served the purpose of visual storytelling, often religious, and for the equivalent of postcards — painting of landscapes, domestic scenes, still lifes and other somewhat utilitarian visual information capture. Painting served the semi-practical purpose of faithfully and accurately capturing scenes for the purpose of sharing.
Imagine the panic among professional painters in the 19th Century. A new technology emerged that could create a better (more “realistic”) image that took less time and, eventually, anyone could do and that could be easily reproduced — someone could make a thousand photographs from a single picture.
But photography didn’t end painting. Instead, photography improved painting in two ways. First, impressionists and other paintings immediately started using photography (often secretly) to improve their work. They were able to freeze in time fast-moving phenomena, such as ocean waves, and really look at them. They were able to study the nature of light, capture exact proportion and use photographs of human models that would hold perfectly still. The painter’s collaboration with photography improved the goal of faithful, accurate representation and ushered in new photography-enhanced painting styles like realism and naturalism.
Second, it freed painters to explore non-faithful reproduction — to capture human emotions more creatively and explore human visual peculiarities. The birth of impressionism was fundamentally counter-photographic. Outdoor scenes were created by blobs of colors — what we might today call “pixels.” Art Noveau, Cubism, Futurism, Surrealism, Abstract Expressionism and all the rest were inconceivable without the faithful representation load being carried by photography.
In addition to vastly improving the painting arts, photography ushered in a whole new art: Photography itself — an expressive tent big enough for gallery-filling artists and billions of amateurs.
While photography initially looked like an existential threat to painting, it turns out that it was the best thing to ever happen to the medium.
How AI is likely to affect human expression
In the new age of easily available generative AI, everyone is concerned that human creators will be replaced by machines, ushering in a dystopia where all our content comes from computers.
But if the photography example is any guide — and I believe that it is — the AI content boom will change everything, and mostly for the better.
With visual art, just like painters used photographs when photography first emerged, AI visual artists (and amateurs) use photographs. It’s already happening with the creation of social media profile pictures.
Photographs radically improve AI images, and AI radically expands what a photographer can do with a photograph.
But the biggest improvement is likely to come with AI that can write, explain things, express itself in any style and craft perfect prose. This brand of AI will confront us with what is special about human-written language and verbal expression. We’ll let the machines write the formulaic dreck, freeing up writers to explore new realms.
The truth about generative AI that writes prose is that its expression is based entirely on what has already been previously written. And so it can only replace the kind of human writing that itself copies previous human writing. It’s an activity unworthy of the human creative mind.
The lesson from students cheating with AI-generated essays, for example, is that humans shouldn’t need to write essays anymore. Academia should and probably will start grading on originality exclusively.
In journalism, most of the articles we read are pickups. Hundreds of news outlets read the original reporting of other publications, and regurgitate it. Let the machines do that. AI will help struggling newspapers, magazines and news websites survive financially.
Human journalists will be freed to do more investigation, more creative explanation and write more and better opinions. (I’ve always felt that good editorialization is undervalued. Of course, I’m an opinion columnist, so — grain of salt.)
We don’t need humans to write bland, basic, derivative, explanatory drivel of the kind that CNET Money so easily replaced with computer-generated articles.
We need to write content that AI cannot write. And AI will force us to do so.
The worst thing we could do is ban AI solely for the purpose of employing humans. The revolution is coming. And we should adapt like we did with the advent of photography — by making content better. And we will.
Mike’s List of Brilliantly Bad Ideas
1. This typewriter writes your novel (so you don’t have to!)
The Ghostwriter, conceived by designer Arvind Sanjeev, is a repurposed Brother typewriter that taps into ChatGPT via an installed Arduino. Instead of typing the words directly on the page, you instead type text prompts for ChatGPT, and the AI sends back the words for the Ghostwriter to type on paper.
2. Gadget won’t let you type “LOL” unless you really did laugh out loud
Brian Moore’s LOL Verifier exists to keep you honest. It stops you from typing LOL if it didn’t actually hear you laugh. When you type LOL, it listens. And if it hears you laughing, it types LOL, plus “verified at 2:34pm” or whenever the laugh took place. Now THAT’s funny.
3. Don’t waste water in the shower. Use it to power your electric toothbrush!
A would-be inventor named Dustin Morgan is Kickstarting a shower head attachment that uses the flowing water in your shower as a kind of hydroelectric dam for the generation of power. Water + electricity? What could go wrong?
Mike’s List of Shameless Self Promotions
Hybrid-work hardware debuts at CES. Wait…, what?
The one real problem with synthetic media
How can the White House’s new IoT labels improve security?
Twitter is the new poster child for failing at compliance
ChatGPT: Finally, an AI chatbot worth talking to
Reframing our understanding of remote work
Where I’m at: Silicon Valley, California
Regarding the shower-powered device, I think it's brilliant. Obviously there isn't enough power generated to hurt someone but it might be just enough to adjust water temperature automatically.