Perhaps because of the visual nature of Biomorph Land, the first
people to incorporate Dawkins's idea of computational breeding were
artists. The first was a fellow Brit, William Latham; later Karl Sims in
Boston would take artificial evolution further.
The exhibited work of William Latham in the early 1980s resembled a
parts catalog from some unfathomable alien contraption. On a wall of
paper, Latham drew a simple form, such as a cone, at the top center, and
then filled the rest of the space with gradually complexifying cone
shapes. Each new shape was generated by rules that Latham had devised.
Thin lines connected one shape to its modified descendant shapes. Often,
multiple variations would split off one form. By the bottom of these
giant pages, the cone forms had metamorphosed into ornate pyramids and
art-deco mounds. The logical structure of the drawing was a family tree,
but with many common cross-marriages. The entire field was packed; it
looked more like a network or circuit.
Latham called this "obsessive, rule-based process" of generating
varieties of forms and selecting certain offspring to develop further,
"FormSynth." Originally he used FormSynth as a tool to brainstorm ideas
for possible sculptures. He would select a particularly pleasing form
lifted from the map of his sketches and then sculpt the intricate shape
in wood or plastic. One of Latham's gallery catalogs shows a modest
black statue with a resemblance to an African mask that Latham created
(or found) using FormSynth. But sculpting was so time-consuming, and in
a way superfluous, that he ceased doing it. What most interested him was
that vast uncharted Library of possible forms. Latham: "My focus shifted
from producing a single sculpture to producing millions of sculptures,
each spawning a further million sculptures. My work of art was now the
whole evolutionary tree of sculptures."
Inspired by an avalanche of dazzling 3-D computer graphics in the U.S.
in the late 1980s, Latham took up computing as a way to automate his
form generation. He collaborated with programmers at an IBM research
station in Hampshire, England. Together they modified a 3-D modeling
program to produce mutant forms. For about a year artist Latham manually
typed in or edited gene values in his shape-generating program to
produce wonderfully complete trees of possible forms. By modifying a
form's code by hand, Latham could search the space at random. With
understatement Latham recalls this manual search as being
In 1986, after encountering the newly published Biomorph program, Latham
merged the heart of Dawkins's evolutionary engine with the sophisticated
skin of his three-dimensional forms. This union birthed the idea of an
evolutionary art program. Latham dubbed his method "the Mutator." The
Mutator functioned almost identically to Dawkins's mutating engine. The
program generated offspring of a current form, each with slight
differences. However, instead of stick figures, Latham's forms were
fleshy and sensual. They popped into one's consciousness in three
dimensions, with shadows. Whole eye-riveting beasts were drummed up by
the hi-octane IBM graphics computer. The artist then selected the best
of the 3-D progeny. That best form became the next parent, begetting
other mutations. Over many generations, the artist would evolve a
completely new three-dimensional body in a true Borgian Library.
Biomorph Land -- huge as it was -- was only a subset of Latham's space.
Echoing Dawkins, Latham states, "I had not anticipated the variety of
sculpture types which my software could create. There appears no limit
to the wealth of different forms that can be created using this method."
The forms Latham retrieved, rendered in mind-boggling detail, include
elaborately woven baskets, marbled giant eggs, double mushroom-things,
twisty antlers from another planet, gourds, fantastical microbial
beasts, starfish gone punk, and a swirling multi-arm Shiva god from
outer space that Latham calls "Mutation Y1."
"A garden of unearthly delights," Latham calls his collection of forms.
Rather than try to imitate the motif of earthly life, Latham is after
alternative organic forms, "something more savage" than life on Earth.
He remembers visiting a county fair and stopping by an artificial
insemination tent and seeing photographs of gigantic mutant superbulls
and other kinds of "useless" freaks. He finds these bizarre forms
The printouts are surrealistically clear, as if photographed in the
vacuum of the moon. Every form possesses a startling organic feel to it.
These things are not copies of nature but natural shapes that do not
exist on Earth. Latham: "The machine gave me freedom to explore forms
which previously had not been accessible to me, as they had been beyond
Deep in the recesses of the Borgian Library, racks of graceful antlers,
shelves of left-handed snails, rows of dwarf flowering trees, and trays
of lady bugs await their first visitor, whether that be nature or
artist. As yet, neither nature nor artist has reached them. They remain
unthought of, unseen, unmaterialized, mere possible forms. As far as we
know, evolution is the only way to reach them.
The Library contains all the forms of life past and life future and
even, perhaps, the shape of life present on other planets. We are
blocked by our own natural prejudices from contemplating these
alternative life forms in any detail. Our minds quickly drift back to
what we know as natural. We can give it a momentary thought, but we balk
at filling in much detail on so whimsical a fantasy. But evolution can
be harnessed to serve as a wild bronco to carry us where we can't go by
ourselves. On this untamed transport we arrive at a place stuffed with
odd bodies, fully imagined (not by us) down to the last hair.
Karl Sims, CM5's artist, told me, "I use evolution for two reasons. One,
to breed things I would have never thought of, nor would have found any
other way. And, two, to create things in great detail that I might have
thought of, but would never have time to draw."
Both Sims and Latham stumbled upon discontinuities in the Library. "You
develop a feel for what kinds of things can happen in an evolutionary
space," Sims claims. He reported that he often would be evolving away,
making satisfactory progress -- sort of whistling happily while things
noticeably improved -- when suddenly he'd hit a wall and the improvements
would plateau. Even drastic choices would not "move" the sluggish form
away from the rut it seemed stuck in. Generation after generation of
progeny seemed to get no better. It was as if he were trapped on a large
local desert basin where one step was identical to the next and the
interesting peaks were far away.
As Thomas Reed stalked the lost chalice in Biomorph Land, he often
needed to back up. He would be near the cup but getting nowhere. He
often saved intermediate forms on his long chase. Once he needed to
retreat hundreds of steps back to the sixth archived form in order to
get out of a dead end.