EndlessForms by Jeff Clune


Talk abstract:
I will describe my work with an algorithm that combines evolution with
developmental biology to automatically produce complex designs. This
algorithm enables the generation of complex, regular phenotypes, such as
three-dimensional objects and artificial neural networks to control
robots. The evolved morphologies and neural networks exhibit desirable
properties found in their natural counterparts, such as symmetries and
repeated motifs. I will also describe a project that combines this
algorithm with crowdsourcing to allow internet users to collectively
explore the evolutionary design space. At EndlessForms.com, visitors
individually direct evolutionary processes similar to how breeders
evolve roses or dogs. Collectively, visitors have evaluated over 2.4
million objects and have produced an interesting and diverse set of
evolved objects.

Jeff Clune is a Postdoctoral Fellow in Hod Lipson’s lab at Cornell
University, funded by a Postdoctoral Research Fellowship in Biology from
the National Science Foundation. He studies generative encodings, which
enhance evolutionary algorithms by augmenting them with concepts from
developmental biology. These concepts enable the assembly of complex
neural networks and other forms from compact genomes. One of Jeff’s
current projects is EndlessForms.com, a website for designing and printing
3D objects with such encodings. Jeff also develops evolutionary algorithms