Evolve Beauty

art, aesthetics, collaboration, evolution

30 July 2009

Garden Update - Going Forward

It has been about three months since the Garden has fully gone live. It has received a lot of attention. In June, I had an open studio where many people who had never heard of the Garden came and witnessed it first hand, both in front of my computer and on the wall in the form of some of my favorite plants which I printed out and framed. I handed out cards with the address of the Garden and received many hits on my web site. Many new plants came into being from the surge in visitor interactions.

But, the current Garden is just version 0.1 of this project. It suffers from some shortcomings. Some of these are related to the mechanisms I developed for the genetic drawing instructions. Some are simple interface issues. Some limitations are inherent in the medium: no matter how cool an idea may be, it is rare that one visitor will make two trips to the same web site, unless rewarded heavily with new information or unless there is a transaction of ownership taking place.

I would like to address some of these shortcomings by redesigning the genetic code used by plants, improving the interface to the Garden and increasing a visitor's sense of ownership.

Currently a plant's genetic code is highly sensitive to small modifications. That is, by changing one gene, there is a cascading effect such that all subsequent genes change also. This worked well in fostering a high degree of variation, but did not work well when it came to make small modifications to an established genome. In fact, this sensitivity pretty much precludes evolution taking place in any coherent way. One of the first and probably most significant changes in the genetic code will be to make genes more concretely independent of where they are located in a plant's genome. This should have the immediate result that visitors will be able to see small variations and a progression from one plant form to the next.

There are many other changes I plan to make in the drawing instructions available to a plant, and I'll be writing more about these as they come up. But, this is not the only area in which the Garden could be improved. The interface is a bit clunky. It's hard to navigate through many plants quickly. It is hard to see how one plant begets another. I made an early decision to simplify the interface, perhaps at the cost of not providing enough information to the casual visitor. (The decision was also made in part due to time limitations.) But it was my fear that too much information might run people off.

However, in my first in-person presentation of this project (at the open studio), I found myself explaining the same complicated, but key concepts over and over. And I was surprised by how much more excitement I could see generated by knowing how the Garden was built, how plants reproduced, what determined their shape, how incidental interactions with visitors greatly influenced the beauty in the resulting plant forms. So, not only do I plan to build a more intuitive interface to the Garden, but I also want there to be more information at every level. I want to include statistics, explorations into plant genomes and more relational information, things like: show me the plants I voted for, show me the genealogy of a particular plant, show me popular plants, etc.

These last points, especially regarding popularity and votes, contribute to a sense of ownership. I believe that if visitors feel they own their contributions to the Garden and can track the progress of the plants they vote for and interact with, they might come back more often. I think only by this mechanism will the Garden truly flourish.

Evolution is a slow process and it takes millions of small events to mold complex and appealing forms. At some point, I think the Garden needs to live in peoples' phones, not just on my web site. People should be able to cultivate their own gardens if they like. To get really interesting, this project needs some years of active collaboration (in many platforms and venues), just like a real garden needs time for plants to become fully established. This is only the beginning. Stay tuned.