Joshua Vermillion has a thoughtful way of saying “nonbiological intelligence” that underlines just how deeply mysterious AI is, while also dialing back the rise-of-Skynet vibes that invariably attend this topic. This isn’t surprising: Vermillion, an associate professor at UNLV’s School of Architecture, explores AI’s uses in the design trades, both in his own nationally recognized creative practice and in the classroom. Thoughtfulness is required in both domains, and the edited-out portions of this interview swell with references to the need to bring critical thinking to bear.
In honor of Desert Companion’s Photo Issue, we asked Vermillion about his Instagram page, where the curious can see his ongoing experiments in image-generation, and his thoughts about AI imagery more broadly.
Let’s imagine a spectrum: At one end is “extreme AI skeptic”; at the other is “AI evangelist.” Where would you locate yourself on that spectrum?
It really does seem like it’s one or the other, right? Alarmists on one end, and then the folks trying to raise billions of dollars to train the next model. So I try to stay in the center. And, really, I think it’s a process of trying to understand what’s signal versus what’s noise. A lot of words get thrown around, oftentimes very carelessly, and there’s a lot of almost propaganda on either end. So, I try to stay in the center, where I am both excited about experimenting with these tools — but at the same time, you have to come at it with a critical point of view, a critical lens, ask some tough questions about it, and that’s what I try to be.
People can get an immediate look at some of your work on your daily Instagram feed, where you post these speculative architectural spaces you generate with AI, each one usually centerpieced by some showstopping design element. What are you trying to do with that project?
My take is, ambassador or skeptic, you need to have an informed opinion about these tools, and so, trying to understand what they can and can’t do became a daily posting. I’m always thinking about, what kind of information do I pump into these AI models, and what can they actually return? They can generate such exquisitely rendered effects, spatial effects, lighting effects, textural and material effects — I’m trying to understand what kinds of things it can actually draw, even though AI has no idea what light or materials or textures even are.
Do you think there’s a concern about people feeling they can’t trust imagery anymore?
That was a big concern with the most recent election, right? The importance of trying to understand what’s fake and what’s real. … And I would argue that’s always been a concern. I mean, technologies can always be used for good or bad, right? That that holds up a mirror to us humans to ask ourselves, How do we use these tools?
I see a lot of recycled, unserious critiques of AI, but there is really room to help ourselves understand what’s innately human, what can’t be automated. Things like ethics and empathy, those qualities that are probably more important now than ever.
What I would hate, though, is to unequivocally dismiss a technology that is really going to be transformative in the long run. I don’t think we can do that.
