Organisms that respond quickly to changing environments have an advantage over those that don’t. However, reacting too quickly wastes time and energy in tracking meaningless environmental changes. Former SFI Postdoctoral Fellow Eddie Lee (Complexity Science Hub), SFI President David C. Krakauer, and former SFI Professor Jessica Flack have devised a mathematical model for optimal learning in a changing environment. In a study published in Proceedings of the Royal Society B: Biological Sciences, they derive a scaling law that shows that an organism’s learning rate should change as the square root of the rate of environmental change.
They also ask what happens when organisms seek to engineer the time scales in their environments — adding structure or erasing structure through niche construction. While niche constructors slow down environmental change, they still obey the scaling law of memory — as long as niche constructors monopolize the niches they create.
Lastly, the researchers looked at how learning and metabolic costs intersect. For small, short-lived animals, learning costs exceed metabolic costs. Conversely, metabolic costs dominate for larger, longer-lived animals, or for those whose environments change slowly, promoting longer memories.
Read the paper "Constructing stability: optimal learning in noisy ecological niches" in Proceedings of the Royal Society B: Biological Sciences (October 30, 2024). DOI: 10.1098/rspb.2024.1606
Read more
Why Elephants Never Forget, but Fleas Have, Well, the Attention Span of a Flea (Complexity Science Hub, October 31, 2024)