Artists and engineers Karl Sims and Alexander Mordvintsev speak about the potential of artificial life (A-Life) and creativity with Peter Bauman (Monk Antony). They discuss A-Life as a bridge between technology and biology, from Sims’s early work with the legendary Connection Machine and Evolved Virtual Creatures to Mordvintsev’s seminal DeepDream.

Alexander Mordvintsev: What are you working on now?

‍Karl Sims: At the moment, I have a couple projects. I've been installing a few museum exhibits with real-time interactive visual effects. I have some pieces at the Museum of Science in Boston, one at MIT, and one at the Museum of Mathematics in New York. Plus I’m installing one at the Boston Children's Hospital. These involve fluid flow, particle systems, reaction diffusion, and so on, simulated in real time so visitors can play around and experiment.

I'm also working on a WebGPU-based application for image synthesis and special effects, replicating some of the types of things I was doing ever since Connection Machine days but now are much easier on the web. That's a longer-term project though so I don't have much specific to say about it yet.

Peter Bauman: Alex brought up your mutual connection to nature and representation, which leads to the very beginning of your story, Karl, because you studied life sciences at MIT. But fascinatingly, you went on to this incredibly forward-thinking organization, Thinking Machines, which is like an early OpenAI or Anthropic today.

‍How did a degree in life sciences lead you to this radically experimental early AI company?‍

‍Karl Sims: Right. Yeah. They weren't hiring a lot of biology majors.

When I was an undergrad, I was also very interested in computer science and computer graphics. I took classes in that as well, even though it wasn't my official major. I also had an internship at the Architecture Machine Group, which later became the MIT Media Lab. So when I graduated, I already had some experience with computer graphics. My roommate was Brewster Kahle, who was at Thinking Machines at the time and went on to found the Internet Archive later on. He introduced me to Danny Hillis and the gang at Thinking Machines and I ended up joining them and trying to figure out how to program this crazy new parallel computer—to make pictures and computer animation with it.

I actually worked at Thinking Machines twice. After a while, I returned to the MIT Media Lab for a graduate degree, then worked in Hollywood for a while, and then returned to Thinking Machines. I was jumping around a little but was actually just following these Connection Machines. The Hollywood firm was an early customer and the Media Lab had also acquired one.