A 'no-brainer' paradigm of collective intelligence: how might non-neural systems generate complex patterns
Zoom link: https://umontpellier-fr.zoom.us/j/93775668851
Abstract
The ability to generate complex patterns is an indication of the capacity for nonlinear spatiotemporal information processing and even intelligence. The biological brain is often touted as a system uniquely capable of such feats but certain non-neuralsystems (e.g., slime molds) have also been known to be endowed with similar capabilities albeit at smaller scales. At the Levin lab, we have been interested in understanding the informational mechanisms underlying the complex behaviors of such non-neural systems.In this talk I'll showcase minimal differentiable models of non-neural dynamical systems, consisting of non-excitable cells with only short-range connections, that generate patterns of varying complexity. Using simple tools from calculus we can reveal the macroscaleinformation-processing mechanisms to better understand how these systems generate complex patterns. Importantly, these high-level but low-dimensional dynamical mechanisms are more interpretable than the descriptions of the corresponding microscale mechanisms(the model) in ways that provide tantalizing hints as to what makes these non-neural systems "intelligent”.
Further information:
Please share and spread the word, the more the merrier!
If you've any other question or request, post in the group or contact us at
leo [dot] zeitler [at] pasteur [dot] fr
gautham [dot] ganesh [at] cbs [dot] cnrs [dot] fr