Been reading a bit about the concept of distributed decision making/computation recently. The essay in question discussed the idea that currently, construction of a complex machine (an aircraft carrier, say) cannot be carried out by one brain working alone, or even a series of individual brains working in isolation of one another. Rather, its necessary to “hook up” a bunch of brains into a interlinked network - engineers, weapons systems designers, logistics managers, construction workers all working and communicating, both with each other in an overarching project framework.
In developing this final product, these hooked up brains can be considered to be a networked series of processors, with the links between each processor being the verbal exchanges, emails, documents and other kinds of informational exchanges between each worker building the aircraft carrier. A networked series of processors is effectively the basis of mind – an obvious parallel being the interlinkages of neurons in the human brain, or chemical exchanges between individual members of an insect hive mind. The more complex the interlinkage, the greater the potential for the emergence of consciousness – compare the complexity and number of interlinkages in a human brain to the number of interlinkages in an insect colony.
The essay went on to say that the power of a group “mind” is limited by the speed at which the informational exchanges between each constituent of the network (each brain) can take place – ie, the speed at which one individual can explain, and another can absorb, pieces of information. As such, you wouldn’t expect significant synergies from such a networked processor – while it will produce the expected final product - an aircraft carrier in this instance - its unlikely that any greater product /output than this will be achieved. That is, its unlikely that anything that could be described as “consciousness” would emerge.
However, increase the speed and complexity of the interlinkages between each brain, and you increase the power of the overall networked processor. Accordingly, you may start to see much more complex interactions emerging – the end result being not just the development of the expected aircraft carrier, but maybe something far greater in complexity and usefulness. This is the concept of emergence – self sustaining complexity emerging from sufficiently complex underlying physical systems. Effectively, what you might see could be the emergence of AI from the interlinkage of individual human brains – a consciousness greater than the sum of the individual minds that make it up.
So – one way to achieve such emergence may be to increase the speed of communicating and processing of the information exchanges between each brain. However, aside from developing new, highly efficient codified languages (faster talking), or directly augmenting the processing capabilities of individual human brains (faster absorption of information) I’m at a loss as to how this could be achieved with current technology.
However, a potential alternative is to consider whether by increasing the scale of the networked minds, rather than the speed of the interlinkages between its constituent parts, we might see a similar increase in processing power. Potentially, with reasonably fast internet connections, and more importantly, increased connectivity of a large number of individuals, you might see the formation of a sufficient number of complex pathways. The question is, are there enough sufficiently complex and sustained pathways to result in the emergence of something which might be called mind?
This concept is particularly attractive as it might be a way to facilitate effective group action. I’ve suggested earlier that the basic failure of centralised or planned economies is the lack of a smart and sufficiently materially disinterested central planners. Could the emergence of a human group mind/AI resolve this issue?
No comments:
Post a Comment