This post is part of a collection of thoughts I’m putting together as a companion to Five Seven Five, a stand-alone book that uses haiku I wrote as input to an AI art generator. It was a super-fun project to build. You can pre-order it at several online retailers.
Fun or not, though, the use of AI in any endeavor is filled with intrigue. So as I came closer to launching the project I decided I wanted to explore the bigger picture of how AI is impacting our world, and specifically the creative world I dwell in.
Today let’s talk about all the things, and the three types of people
I really did intend this series to be an everyday thing during the span of Kickstarting Five Seven Five, but life is like that. Being open, we’re finalizing our move process to Las Vegas and I had a couple day blip that derailed a lot of my schedule.
Go figure.
Anyway, I did want to spend some time discussing the overall situation as I see it, now that I’ve completed a project that dabbles in the world of AI.
My view today: As I said in my last post, I think we have a human problem—not an AI problem. They are real problems, of course. But they are problems in our minds and our lives, and not specifically with the AI itself. In other words, they are problems associated with adjustments we need to make to live in a world with AI.
The AI being invented here is not capable of thought. Of course it’s not. Hence, it’s not capable of conveying a purposeful message. In this way, an AI is a meat and potatoes tool rather than an entity in itself. That’s maybe why it’s gotten a foothold into the world of journalism, which is often viewed (poorly) as a “just the facts, ma’am” kind of world. If a person knows the who, what, why, when and how of anything that happened—that person can turn out a rapidly written “professional” piece almost by rote. So can an AI. Duh.
That is just one part of journalism, however. And it’s a part a lot of us wail about in poorly made arguments in which we pretend that’s all we want—just facts—when, in reality, we want stories, too. Because stories give us something beyond the five questions. Stories give us perspectives and context. Stories help us decide how we feel about things. An AI journalist simply can’t do that today, and it’s arguable whether it ever will.
(Note, the good faith argument that can be made that an AI will eventually be able to supply human context goes like this: If an AI can be trained on all pieces of data that exist, regardless of copyright, then that AI probably can be considered to know more about human life than actual humans do—because the AI is not constrained to view the world from the lens of only one human. Regardless, that’s party discussion fodder for now.)
Despite the fact that AI are unlikely to be able to match a human in creating what it means to be human—there exist a weird set of arguments and counter arguments about how AI will kill various forms of art. These debates are tautologies that literally cannot be true. If AI cannot replace artists, then it cannot kill art. Alternately, if an AI kills art, then by definition it must be capable of replacing artists.
Having done this project and dwelled in the space a while, I fall into a group that thinks the arguments being waged are simply false ones. I think AI art and human art will coexist, and I think the human problems associated with this middle zone will be solved by deciding how we are going to segment and present AI assisted mechanisms related to creation to the world at large.
An example: As I move into the audio book space (which I played with last year, and intend to get into more deeply this year), I can see releasing three different audio books: one read by me, one read by a human I’ve hired, and one read by an AI. The pricing would be different, and the projected market for listenership related to each would be different, too. Of course, there would be commercial constraints, which might affect both the schedule and the choice to actually do something. The AI (on Google), is free, but it takes a little time. Since I can do a bit of a decent job on voicing and recording my own work, the use of me as a narrator is also free regarding currency outlay, but is a huge time expense to record, edit, and distribute the end product. My ability to hire humans is limited by cost. And, if I don’t have money to fuel the project, the question is not “human or AI,” the question is “audio book or no audio book?”
So, yes, I think there are niches for each part of the spectrum. In the audiobook case I mentioned, the main requirement is that I properly mark each so that the listener knows what they are buying. Audio books are cleaner, though, because the voices that have been used are humans who have been paid to train the AI.
In doing this project I’ve come to the edgy conclusion that there are three kinds of people in the idea space around the use of AI in creative pursuits right now.
Type 1 are folks who are simply offended by the existence of the AI to begin with. They worry that art is coming to an end, and they are certain that use of material for training the AI is theft. They also focus on the AI’s ability to create something close to replicas of originals as if that’s something dangerous and new. Perhaps they are right on all counts, but I obviously don’t think so.
Type 2 are folks who don’t really want to be bothered, but aren’t getting caught up in the handwringing. They are busy doing what they’re doing, and don’t want to get bogged down. I think of these as the “true artists,” and in my mind they are the ones who will keep working in their own ways until simply forced to change (it strikes me to compare them to film photographers, who stuck with their darkrooms and chemical processing as digital photography bloomed around them).
Type 3 are folks who decide that AI tools can help them. Perhaps I should further segment this group, but for now I won’t. There is, however, a broad spectrum here. In my case, I don’t expect to roll AI into my “real work” for now, though I’ll be looking at it closely. Five Seven Five was a fun project, purposefully geared to explore AI, but I’m not so thrilled with it to use it in my other projects. Maybe. I don’t know, really. Since I’ve been doing my own covers for a bit now, I’m probably more open to using AI art in spots than AI text creation. And since I don’t have an audio book presence I’m more open to the less expensive and less time intensive approaches to AI there. As I put it in the accompanying video to this piece, maybe I fall into a type 2.5. Whatever. I’ll suggest that if AI-driven audio succeeds for me, it’s much more likely that I’ll eventually be able to afford to break into hiring humans to do top end work—so just for me as a creative, the existence of AI as a platform may well result in more money flowing to a human being than would have without it.
Fuel for thought, anyway.
Regardless, the main point I’ve come to in working on this project is that I know I don’t want to be a type 1 creative, and that I’m probably not destined to be a type 2 in the end. How far I go into type 3 is a question for the future.