How Do We Create, Define, and Search for Life?

facebooktwitterreddit

Christoph Adami makes "artificial life," effectively self-replicating computer programs. If you've heard of Tierra*, that's the kind of thing Adami does. In this TED Talk, Adami discusses his work, including a brief discussion of how we define life on Earth -- which is lots of fun, as he describes a real-world organism that does not die. And no, we haven't been overrun by this monster (yet).

After the initial discussion of "what is life" (in the context of wanting to identify extraterrestrial life), Adami explains his own artificial life programs, and that's when things get deeply geeky. If you're interested in science of any kind, this talk is worth your twenty minutes. Stick around for the animations showing biodiversity, and then the animation showing how the rate of mutation affects populations -- it's strikes a very weird geek chord to see a fellow geek demonstrate his "it's alive!" moment onstage.

* = Brief personal anecdote on artificial life: in high school, I became interested in Tierra and similar systems. So for my high school's science fair circa tenth grade, I wrote a life simulator that attempted to demonstrate a form of "natural selection" (well, non-natural, but at least selection) based on mutation and competition within an artificial landscape. My project was a C program that created various artificial life forms (predators and prey) that competed in a virtual landscape, reproduced, mutated using an ultra-simplified DNA-ish structure, ate each other, died of old age, and so on. The program's output was text-only, indicating the genomes and population counts of dominant organisms. The program was so resource-intensive that I had to borrow computers to run it on, in order to actually demonstrate the system's long-term viability -- my home PC was so slow that a single "generation" took several minutes to run, and I needed to run thousands or millions to demonstrate the long-term effects of small mutations. Anyway, I won the Computer Science division of the fair (by virtue of being its only entrant), but didn't place overall. Oh well -- I learned that it's really hard to explain "artificial life" to non-geeks, so it's encouraging to see Adami's presentation, particularly his clever animation showing the mutation threshold necessary to sustain artificial life.

You might also enjoy a photo of me with an elementary school science fair project. I was working with real, plant-based life at the time.

(Via Hypercritical episode 67.)