Intelligence is the capacity of a complex system to flexibly modify its activity toward specific goals states. Much depends upon the ability to detect intelligence, even at drastically unexpected time and length scales. Discussions of which animals have intelligence—and the near-ubiquitous focus on the entire animal as a single, privileged level of focus for intelligence—only address the tip of the iceberg.

Rather than posit intelligence as a characteristic of whole organisms, Michael Levin hypothesizes that it is present at multiple levels and scales within every living creature. From the subcellular to organs, and from individuals to swarms and collectives, his team will explore these nested intelligences. This groundbreaking study seeks to broaden our sense of intelligences everywhere.

Recent efforts in the science of primitive cognition, the philosophical direction of panpsychism, and mathematics (e.g., Integrated Information Theory) have expanded our definitions of intelligence in much-needed directions. But one limitation is that it remains unclear how to move beyond philosophy to empirical tests of these ideas.

This project will establish a proof-of-principle for using expanded definitions of intelligence to reach new capabilities at the bench. The team will go in two directions, above and below the individual:
(1) train an ant colony (not individual ants) via instrumental behavior shaping, demonstrating communication with a group mind as empirical evidence that a group intelligence exists in this system;
(2) produce trainable biological systems comprised of embryonic cells and tissues, showing that the components of the body are intelligent.

Outputs will include scientific papers and presentations at conferences. The team will also create an intelligence detector kit: a simple set of protocols for disseminating these ideas into the public (from high-school to post-doc levels).