Diving in with AI
What's been going on in the lab?
I've been catching up and finally diving into AI. While I've had my reservations about it - primarily because there are those among us with selfish and nefarious intentions - it's become nearly impossible to ignore the impact that AI is already having on our culture.
About a month ago, I signed up for Midjourney to see what it was all about.
It took a bit to get into the flow and figure out how I would use this tool, but a few days in I was having a blast with it. In the past 30+ days, I've generated hundreds of images. Here are my takeaways thus far:
1. AI as a collaborative tool.
There's something interesting about seeing it as a collaborative tool, where I, the human, am collaborating with the machine. There is some fear about how much AI will ultimately replace the functionality of humans. While I both understand and also harbor some of that fear myself, I think there is something unique and interesting about its current state whereby it requires human input. I, the human, have to provide it with prompts for it to perform its function. By that I mean, that I, a person with a unique point of view and set of experiences, am inputting the prompt and then making decisions about which output best represents that point of view. Which leads to my next takeaway.
2. AI as a tool for curation.
Rather than using AI image generation as a tool to replace traditional art and design making, I like seeing it as a tool for curation. As a music fan and general collector of things, I've always loved the act of “digging,” a term often ascribed to searching through bins of vinyl to find unique gems. When my perspective of AI shifted to this mindset, it freed me up of any of those preconceptions about it replacing creation and shifted more to a mindset of discovery. Which is a nice segue into takeaway 3.
3. The power of conceptual prompts.
I see a lot of people using tools like Midjourney, DALL-E, and Stable Diffusion to get very specific results. I too have used it in such a way and have gotten some interesting and very impressive results. However, it's when I've input less specific visual prompts and used more interpretive language that I've created some of my favorite images thus far. I am someone who has always drawn to more conceptual and abstract art, art that often asks that the viewer to participate in determining the meaning. In this use case, there's almost a reversal of roles. I am inputting an idea and the creator, in this case, the machine, is interpreting the idea. When the results are given, I am given the task of both curating the results and in many cases, closely examining the results and finding even deeper and adjacent meaning.
Please enjoy this small sampling of the images I've generated in the last few months.
While I still have a lot to learn, I have found that this time of experimentation has transformed my initial perspective on AI, and in particular, text to image generation.
And although I am still skeptical about how humans could use these and other AI tools in ways that could cause harm and confusion, and stoke division, I am trying to remain hopeful and believe in the goodness of people. Yes, this is me attempting to look past my own cynicism and lean into optimism, but the truth is that with AI, the cat has been let out of the bag and there is no putting it back. It feels like we are in a very transformational time in human history and each of us has the choice to lean in and learn in the hopes of using these tools for good, or ultimately get “left behind” in a way.
I have many other thoughts about technology, the computers and global connection we carry in our pockets, human behavior, as well as a deep nostalgia for a simpler time, but I’ll leave those for another time.
Stay Tuned, there's much more to come!