Design of resilient, adaptive, autonomous robotics will greatly benefit from importing deep ideas [contained in] biological software that exploits noise, competition, cooperation, goal-directedness and multi-scale competency.
ICDO research will build on recent advances from the Bongard and Levin labs, in the following theme/interest areas.
Shaping an emerging field
As we learn more about the deep principles of biological regulation, and the possibilities of machine behavior and machine learning, it is imperative that we continue to examine the ethical and philosophical issues inherent in our work, and build the nuanced language and conceptual frameworks that the ICDO, and all researchers in these fields, need to ground and inspire their work. Many of the core questions and themes of this work are described in a recent publication in Science: Robotics.
AI-driven design of new life forms
[A]n inclusive, continuous view of life and machines frees the creative capacity of bioengineers, providing a much richer option space for the creation of novel biological systems via guided self-assembly.
Bongard and Levin
The announcement of the first computer-designed organism in 2020 was the beginning of an exciting body of multidisciplinary research that the ICDO will expand and deepen in the months and years ahead. Subsequent work will seek to further demonstrate and test the methods proofed out in the first demonstration. Questions driving this work include: what shapes, tasks and goal-driven behaviors can be created in these biobots? What lessons can we derive from complex biological structures and decision-making processes that can inform machine learning, and vice versa? What can biobots teach us about life and intelligence, beyond what we can learn from studying naturally evolved organisms?