Liquid brains are the distributed and dynamic ways of information processing in social insect colonies, slime molds, bird flocks, fish schools, and even the immune system. Unlike solid brains, which are fixed in size and shape, liquid brains can change their volume and form depending on the situation. Liquid brains are composed of individuals that act as neurons, transmitting signals through their interactions with each other and the environment. Liquid brains can also store and retrieve memories, coordinate tasks, learn from experience, and adapt to changing conditions without a central leader or a centralized sophisticated brain.
Solid brains, such as the human brain, exhibit what is termed a small-world architecture. That is most connections are local in nature, while some are not, and these long-range connections carry the group consensus to other portions of the brain. These long-range nerve tracts the so-called rich club are the most active in the brain, and they are particularly good for quick solutions to problems. But if the quick response is met with negative feedback, then a more elegant solution is required and searching among the local small-world component neurons takes place until a new more accurate consensus is reached. The same phenomenon occurs with liquid brains. In bees for example, there is a rich club of individual bees who are termed elites that function at a significantly higher level than the average hive mate and it is this rich club of bees that serve to best find food sources and potential new hive sites, which are then searched by local small-world bees that follow the elite’s instructions that are dictated by a waggle dance (see below).
Decisions are ultimately made by quorum sensing, an emergent phenomenon of greater intelligence than that of the individual bees. Quorum sensing is a democratic process in which individual agents, such as bacteria, fungi, insects, and even neurons in monkey brains communicate with each other and coordinate their actions based on the number and quality of signals they receive from their peers until a quorum level of consensus (say 80% in agreement) is reached. Once the consensus has been reached the group takes the appropriate action. It is an amazing example of how animals can achieve complex and intelligent outcomes without a central authority or leader. By using simple rules and local interactions, they can solve problems that require collective wisdom and consensus. Quorum sensing also shows how nature can inspire new solutions for human challenges, such as distributed computing, network optimization, and social coordination.
You might be asking yourself, “Don’t social insects like bees and ants have Queens?” They do, but they are not part of the executive functions such as decision making, instead they are relegated to mating and egg laying, for which they may lay up to 2 million eggs in their lifetime in some species.
Liquid brains are themselves an example of an emergent phenomenon where the capabilities of the whole are greater than the sum of its parts. Emergent phenomena such as the development of a multicellular embryo proceeds without a leader, an executive, and the embryo as a whole is also greater than the sum of its parts. Thus, emergence can be defined as a Gestalt that leads to a dramatic decrease in description length. To paraphrase Sean Carroll the uncountable number of quantum interactions going on in front of me can be reduced into a single word a table.
One way to understand the intelligence of liquid brains is the ability of a group of individuals to solve problems that are beyond the capabilities of a single individual. These emergent properties are self-organized and robust. They rely on positive and negative feedback mechanisms that regulate its dynamics. For example, if a food source is discovered by a scout bee, it will return to the hive and perform a waggle dance that indicates the direction and distance of the food. The vigor and length of the dance indicates the quality of the food source. The more bees that visit the food source, the more they will recruit others, creating a positive feedback loop that amplifies the signal. However, if the food source becomes depleted or less attractive, the bees will stop visiting it and stop recruiting others, creating a negative feedback loop that dampens the signal. This way, the group as a whole can allocate its resources efficiently and avoid wasting time and energy on unprofitable options.
Finally, one of the most striking examples of how a liquid brain functions is the process of nest-site selection in honeybees. When a colony needs to relocate to a new nest site, such as when it swarms or when its current nest is destroyed, it faces a complex decision problem that involves multiple nest criteria, such as the size, shape, orientation, entrance size, cavity volume, insulation, etc. First, a small fraction of the colony (about 5%, elites), act as scouts that search for potential nest sites in the surrounding area. Each scout evaluates a site based on its own criteria and preferences and returns to the swarm cluster to report its findings. The scout performs a waggle dance that encodes the quality and location of the site. Second, each scout that visits a site compares it with other sites that it has visited before, by the amount and quality of the waggle dancing that is currently going on for the alternative sites. Finally, as more scouts visit more sites and exchange more information, a collective preference emerges among them. Eventually, one site will gain enough support from enough scouts to reach a quorum sensing threshold that triggers a decision. The scouts that have agreed on this site will then stop dancing and start producing a piping sound that signals their readiness to move. The piping sound will spread throughout the swarm cluster and induce the other bees to follow the elites to their new home.
Liquid brains challenge our conventional notions of intelligence and cognition, showing that complex behaviors can emerge from simple interactions among many agents. Liquid brains are truly amazing in their abilities, without a central leader, to perform higher-level cognitive functions, such as decision making, memory storage and retrieval, asset allocation, learning from experience, and adapting to novel stimuli and environments.
Further reading:
Stephen Buchmann (2023) What a Bee Knows. Island Press.
György Buzsáki (2019). The Brain from Inside Out. Oxford University Press.
Seeley, T.D., Visscher, P.K., Schlegel, T., Hogan, P.M., Franks, N.R., & Marshall, J.A.R. (2007). Stop signals provide cross inhibition in collective decision-making by honeybee swarms. Science, 335(6064).
Solé R, Moses M, Forrest S. (2019). Liquid brains, solid brains. Philos Trans R Soc Lond B Biol Sci. Jun 10; 374(1774).
No comments:
Post a Comment