Thinking about AI.

I've been thinking about thinking, and especially when it comes to working with AI. I realized that the way we evolve thinking and AI is quite different. When we are  born, we start crawling around the floor, looking at all the giants around us, looking at images, sounds, grabbing things, tasting them, and noticing that these large human giants seem to be making noises. At a certain point, you realize they're communicating with each other, and you start learning how to mimic them, and then that becomes language.  This takes a while, and that's where the big difference between US and AI. We learn by observing the universe, by seeing, hearing, touching, tasting, and having experiences with other creatures. Language came after intelligence, but AI is the flip side of this. It's learning from language, which is what we use to describe the world after the fact. As beautiful and lovely as language is, it's at best our attempt to describe what we are experiencing in our minds and our brains.

That's one of the biggest differences in how we think and how AI thinks. The disadvantage that AI has now is that it doesn't experience the world as we do, not yet anyway. It's starting to view and design images, but those senses don't exist 24/7. It isn't learning from the world’s visual clues, or from hearing it, or from having direct experiences. It's sort of seeing it after the fact, and that changes how they evolve, but that's a topic for another day.

What I really wanted to talk a little bit about was how I'm thinking with AI as I work on code, films, or animations. I realized at a certain point that what I thought was thinking on my part when I programmed in a traditional way was really just solving technical problems, memorizing how syntax or how you would explain a particular if statement or complex algorithm. Those are the things that AI is pretty good at, and that's what it does well for us.

As I did more coding using agentic programming, which we also called vibe coding, it became quite clear that my efforts to develop something were changing. It was much more about thinking through the process and what I was trying to do, and not really concerned with how to technically achieve it, though that is something you have to keep up with, but it wasn't my main concern. I knew I had to do a lot more testing after the fact, but I also spent a lot of time thinking about where I was going with this, and the experiences I was having kept evolving into new ideas or new ways to solve problems. It was feeling much more like how I would feel if I were painting a landscape.

I was thinking a lot harder than I used to, because now I can sort of think about the bigger picture, the experience, less about the technology of how to implement them. Knowing the technology is exceedingly important, especially if you want to keep AI on the right track, but the real effort comes into how you develop something. As an artist, this became much more the way I would work on a traditional painting or digital art, on game development, on writing a story, or on filming. It was much more about what I was trying to produce at the end, less about the technology or techniques of a brush stroke or of a particular filter that I wanted to use. As Mr. Ross would say, I started discovering “happy little accidents” that led me down interesting paths to addressing a problem or discovering a new solution.

I realized that's where our real power lies, in our ability to do that kind of thinking and worry less about the technical underpinning of everything. As important as that is, the bottom line is we need to understand what we're trying to achieve, and some of that comes in a different way than we used to think about coding. We used to do a waterfall process, where you have well-specified start and end results, usually based on things we knew really well. Then the Agile process came along, allowing us to be a bit more organic, with much more discovery, iterative problem-solving, and letting the problem evolve more naturally. Even that has its limitations, and now, working with AI and team members, both artificial and real, it becomes much more about finding, fine-tuning, and discovering things along the way that you can do better than you needed to do at step three. That could happen either before or later, and that only happens as you go through the creative process of making something. It becomes that you are really thinking a lot more about those important parts of this process and less so about the minutiae of how you put together certain syntax to solve a technical issue.

That has been the biggest surprise to me. I just didn't realize how much I had to work on making something actually good, rather than just fixing a particular call or parameter, and focusing on what I was trying to achieve for the humans using my apps. It feels so good to be creating something again and not just solving technical issues but discovering beautiful solutions.

We can only do so much in a lifetime on a particular job, project, or goal. So we need to focus on what satisfies us most by coming up with solutions that people truly feel and understand, and that elegantly solve problems. I don't miss doing micro-technical research to solve syntax issues. I want to move on to do something truly interesting, truly human, and I think that's what an AI partnership is going to allow us to do: be creative again.