Bacteria Found to Store Memories Like Artificial Neural Networks Without a Brain
Single-celled bacteria have been found to learn from experiences and store the past via ribosomes using a logic similar to artificial neural networks.
A new study published in the journal PRX Life revealed that learning is possible without a brain, and that even a single bacterium can store past experiences to prepare for the future.
Bacteria Adaptation to Nutrient Changes
Researchers tracked tens of thousands of E. coli cells using a microfluidic device as their nutrient sources were altered.
Bacteria exposed to fluctuating nutrient environments were observed to adapt much faster compared to those in stable environments.
The Role of Ribosomes as Memory Storage
A mathematical model of the internal molecular network showed that ribosomes, the protein-producing factories of cells, can serve as a memory center.
While some ribosomes respond quickly to nutrient changes, others change more slowly, providing a memory lasting from minutes to hours.
Similarities with Artificial Intelligence and Neural Networks
This molecular system follows the exact same basic computational logic used in gated recurrent neural networks in artificial intelligence.
Bacteria manage the balance of retaining and discarding information across multiple time scales via molecular gates without the need for neurons.
Potential Impacts in Biology and Medicine
These findings offer a fresh perspective to the world of biology while inspiring artificial intelligence designs for energy-constrained continuous learning.
It is also noted that they hold potential implications in the medical field regarding disrupting pathogenic adaptation.