Cooperation and the evolution of intelligence
April 13, 2012
The emergence of intelligent strategies. Shown are the dynamics during 10,000 generation subsets of simulations for the prisoner's dilemma and snowdrift games. (Credit: Luke McNally, Sam P. Brown, and Andrew L. Jackson/Proc. R. Soc. B)
Trinity College researchers have constructed an artificial neural network model that demonstrates that human intelligence evolved from the need for social teamwork.
The high levels of intelligence seen in humans, other primates, certain cetaceans, and birds remain a major puzzle for evolutionary biologists, anthropologists and psychologists.
It has long been held that social interactions provide the selection pressures necessary for the evolution of advanced cognitive abilities (the “social intelligence hypothesis”), and in recent years, decision-making in the context of cooperative social interactions has been conjectured to be of particular importance.
Study details
The researchers used an artificial neural network model to show that selection for efficient decision-making in cooperative dilemmas can give rise to selection pressures for greater cognitive abilities, and that intelligent strategies can themselves select for greater intelligence, leading to a Machiavellian arms race.
The results provide mechanistic support for the social intelligence hypothesis, highlight the potential importance of cooperative behavior in the evolution of intelligence.
The neural networks (“brains”) participated in two classic social dilemmas: the iterated prisoner’s dilemma (IPD) and the iterated snowdrift game (ISD). In both games, two players must choose between cooperation and defection during repeated rounds. Upon completion of either game, each “brain” produced “offspring” asexually, with “brains” that made more advantageous choices during the games programmed to have a better chance to reproduce.
A potential random mutation during each generation changed the the “brain’s” structure, number of neurons, or the strengths of the connections between those artificial neurons simulating the evolution of the social brain. After 50,000 generations, the model showed that as cooperation increased, so did the intelligence of the programmed brains.
Ref.: Luke McNally, Sam P. Brown, and Andrew L. Jackson, Cooperation and the evolution of intelligence, Proc. R. Soc. B, 2012 [DOI: 10.1098/rspb.2012.0206] (open access)
Comments (7)
by NakedApe
To reinforce the idea of sexual selection, particularly the female’s selection of the male for his intelligence, let me suggest that societies which give greater reproductive freedom to women give rise to more intelligent males. Conversely, societies which oppress women through, say, genital mutilation or stoning for sexual misbehaviour, give rise to very stupid men. The women in such societies just don’t have the freedom to choose their mates. Since intelligence is also transferred between the sexes through genetic recombination, it will also percolate to increase women’s intelligence and thus the population as a whole. Have you noticed that societies which give greater freedom to women are more productive and more enjoyable to live in while societies where women are brutalized tend to have men who are given to superstition, brutality and stupidity? This is just my hypothesis but could be tested by some social scientists or anthropologists, I suppose.
by GatorALLin
…this article had me thinking about the Borg from star trek
http://en.wikipedia.org/wiki/Borg_(Star_Trek)
by eldras
as this has been demonstrated these things now follow:
1. That human brain simulation may best be done as groups and therefore require more processing power than we’ve planned for.
2. That one human brain is part of a network and almost meaningless alone.
The latter means to construct artificial intelligence necessitates group interaction for which much of the brain may exist. Reverse engineering is not of course the only way of constructing superintelligent systems, many of which will arrive by intelligence amplification as Verner Vinge’s NASA paper predicted in 1993.
A caveat is that Hammerhoff-Penrose quantum brain: although quantum gravity interaction at the synapse seem against ockham’s law, such actions have now been found in plants. If consciousness requires these the order of complexity is factored up and may require quantum computers to replicate it.
by Lord Penguin
The brain does not compute with quantum data (the electrical impulses are either 0 or 1 at any given moment), so any quantum interaction at the synapse could be modeled by a probabilistic system.
That’s not to say quantum computers wouldn’t be very useful in creating AI as intelligent as, or more intelligent than humans.
by Jonathan Cole
I see evidence that individuals who are greater risk takers have an edge in procreation. Does it seem that a little bit of ‘crazy’ makes a potential mate more desireable?
by Rob Fleming
OK, the world is complex and everything is connected. But I would suggest that cooperation is a response to selection pressure that becomes self-reinforcing. EG, the more the organism selects cooperation over, say, bigger teeth, the more it needs to cooperate to survive and those that don’t cooperate don’t pass on their genes. This leaves us hard-wired for social interaction, which may explain why social media can become as addictive as alcohol and drugs. So environmental pressure and the response to it are linked, but let’s not confuse the two.
by Kelsey473
It seems to answer why many apes etc would develop `intelligence` but not why one ape `us` is so so far ahead of the rest.
Best guess I would say it Miller`s `Sexual Selection` – The Mating Mind – as to why we make the `extra` jump once the ground work is in place.