3 ms·
This study is specifically about "learning" that takes place without interacting with the brain. It's learning in the same sense the immune system learns to fi
by ants_everywhere 2y ago
This study is specifically about "learning" that takes place without interacting with the brain.
It's learning in the same sense the immune system learns to fight of infections. The difference is that the mechanism by which cells record state is similar to one of the mechanisms also used by the brain at the cellular level, which you would expect.
The cells and structures that make up the brain evolved from simpler structures, so we would expect some reuse of mechanism.
- lupire 2y agoThe obvious-in-retrospect distinction is that memory is everywhere (even a footprintnin dirt is a memory), and selective (reinforced) memory is more complex (as in the OP example), and the combination and communication of memories and information is where the brain really outshines other objects.
- Jerrrrrrry 2y agoUsing the immune system as an analogy as another complexity layer of adversarial <-> learning/adaptation in relation to the context of both complexity theory / machine learning is tempting but almost is too easy. Plus the immune system is likely more complex than the analogy would hold up to.
- akira2501 2y agoThe immune system is several systems. One of my favorite parts is the Thymus. It's literally setup like a gauntlet. New immune cells from marrow come through the Thymus and are tested. They need to pass through and attack foreign cells then pass through and _not_ attack host cells. They are essentially tagged and filtered by this process. Most cells are let go into the body, some cells are reserved as self regulatory cells, and the others that do not pass are destroyed. It's a literal quality control and selection machine for your immune system.