an AI requires apparatus to perceive and act within the world, as well as the ability to move and grow if we want it to eventually exert influence over the entire grid. Most constructions within Life are extremely sensitive to perturbations.
This may be more tractable in Lenia, because lenia’s smoothness means that cell-level changes take many simulation steps and can occur over multiple spatial frequencies. That might also make it harder to study, but I think more people should know about smoothlife variants. Smoothlife gliders have:
free motion relative to other gliders
smooth state spaces
complex phenomena that can be abstracted to simpler ones, a key difficulty of real physics we’d like to have a testbed for
the potential to be trained with gradient descent, allowing generation of very complex patterns to test algorithms such as discovering agency
there’s already quite a bit of work on these topics, check out the page!
This may be more tractable in Lenia, because lenia’s smoothness means that cell-level changes take many simulation steps and can occur over multiple spatial frequencies. That might also make it harder to study, but I think more people should know about smoothlife variants. Smoothlife gliders have:
free motion relative to other gliders
smooth state spaces
complex phenomena that can be abstracted to simpler ones, a key difficulty of real physics we’d like to have a testbed for
the potential to be trained with gradient descent, allowing generation of very complex patterns to test algorithms such as discovering agency
there’s already quite a bit of work on these topics, check out the page!
Interesting. Thank you for the pointer.
The real question, though, is whether it is possible within our physics.