Intelligence and Collective Intelligence

In his book, Superintelligence, Nick Bostrom discusses the likely scenario in which artificial superintelligence develops in the next century. He explores a number of ways in which this could happen, and concludes that they all point to an inevitable intelligence explosion in the near future.

One possibility that Bostrom does not directly address, however, is that different forms of intelligence will tend to undermine each other. Specifically, two types of intelligence he considers, individual and collective intelligence, may work against each other in pursuit of greater intelligence.

Consider the human brain, a highly intelligent collective, but composed of low-intelligence individuals (neurons). Why wouldn’t evolution have produced significantly more complex neurons, if more intelligent neurons would lead to higher collective intelligence of the brain? One possibility is that intelligent individuals actually tend to undermine the intelligence of the collective.

This makes sense intuitively. More intelligent neurons would be less efficient. For example, if they could understand and interpret a greater diversity of signals, each signal would need to carry more information to distinguish it and remove more ambiguity; a neuron that can interpret (2^n) different signals needs at least (n) bits of information per signal.

We also know that in groups of humans, individual intelligence does not correlate with group intelligence. Instead, it is the emotional intelligence of the group which strongly correlates with group performance, which stands to reason, as emotion is an efficient mode of human social communication.

As such, the collective, if it seeks to augment its own intelligence, will suppress individual intelligence in pursuit of that aim. Meanwhile the individuals, as they seek to augment their own intelligence, will undermine the intelligence of the collective. These countervailing forces could tend to slow or prevent the rise of a true superintelligence, or lead to cycles of in which collective and individual intelligence trade off which is improving and which decreasing. This dynamic seems important to Bostrom’s position that signs point to the impending rise of some type of superintelligence.

Evolution fills the cavity - size of skull = size of brain. It only produces what it needs for the given species, and no excess to that.
With brains, it may be that bigger is more cumbersome, so you don’t get 7ft Einstein’s perhaps? though computers and artificial neurons wont be so limited.

My concern with these theories is that they rely upon competition, where I think we will arrive at technology which can build anything at the push of a button [like what’s in my head]. At first there will be licenses on everything [e.g. rolls royce wont be letting people build their cars without a licence to do that], but eventually virtually everything will be bespoke ~ especially as it is as easy to do that as replicate. How can anyone determine rights to shape? Ergo after a while the limits upon design will not make any sense and will be removed. a social ‘revolution’ will occur as people get evermore frustrated if e.g. you cant design anything without every part of that being accounted for in someone else’s design.

Intelligence, refers to Memorization, Analysis and Synthesis of Patterns.

There are two types of Genius.
The realist - A mechanical genius who notices and memorizes patterns. He can fix any car, program any code, build any building without it falling over.
His brain consists of a network of memories of mechanical observations and animations.

There are two subcategories of the realist-genius:
A book-smart realist is someone who’s mental networks are mostly word-based…ie. A mathematician who relies on written laws and principles.
A animative realist is someone who’s mental networks are mostly loose and based on animations.
Both types can fail, a book smart realist can fail by sticking to dogma too much and not having any “common sense”, whilst an animative-realist can fail by having a couple animations remembered differently from what they actually were, so it is best that they work in tandem.

The second kind of Genius is the Abstract genius. It is the opposite of the Realist-Genius, the Abstract genius strives to think outside the realms of physics and mundane phenomenon.

Carleas wrote

Intelligence does not fluctuate significantly, only understanding does. Doesn’t science say that a person uses less than 10% of their brain? How does an individual augment their intelligence?

While I think your first sentence is not quite right (the cavity and the filling evolved in parallel), I think you are right in your general point: there are many evolutionary pressures on the size and composition of the brain, so I shouldn’t rely too heavily on the evolution of the human brain to make my point.

But I provide two additional reasons to think that intelligent individuals are at odds with intelligence groups. One is empirical: studies of human group intelligence indicate that smart individuals don’t make for smart groups. The other is logical/mathematical: smarter individuals make communication more difficult and costly (by increasing the amount of information that needs to be transmitted in order to remove more ambiguity).

So there are three points to support the idea that individual intelligence and collective intelligence are in tension: 1) the logical/mathematical point that having more possible messages means needing more information to distinguish between them; 2) the observation that contemporary attempts to make intelligent groups show that smarter individuals make for dumber groups; and 3) the observation that evolutionary processes have similarly seemed to trade individual intelligence for collective intelligence.

Trixie, I would define intelligence as the ability to solve novel problems. That’s only slightly more precisely defined than just saying “intelligence”, but it is useful and is used in practice by scientists studying animal intelligence. Do your two-types map on to that, or is it a different paradigm completely? Maybe my definition could further define the types of novel problems that can be solved, and thereby distinguish intelligence into realist and abstract?

WD, the 10% figure is false, we definitely use 100% of our brains. There are ways you can word it to make it work (I think it was originally based on the claim that we only understand what 10% of the brain is doing. It may also be correct that only a small number of neurons are actively firing at a given moment, or that most of our brain is axons or some other arguably lesser tissue such that we’re doing the thinking with some smaller percent. But in the sense the claim seems to be interpreted and applied (e.g., the movie Lucy), it is false.

As for augmenting intelligence: in this thread, I mean it in the context of the rise of superintelligence described in Bostrom’s book, where increasingly intelligent machines are able to alter their own code to increase their intelligence. The same notion can apply to humans as well: smart drugs and neural implants are the most direct analog (though also currently science fiction), but education does the same thing on a more modest scale.

And I suppose this too requires us to define intelligence. If it is the ability to solve novel problems, it does indeed fluctuate significantly.

I think the question of the relationship between individual and collective intelligence has to be qualified in terms of stationary-temporal predicates.

In immanent terms of quantifying the variables, as You had, Carleas, it is indeed a fact, that such patent relationship is demonstratively diminishing upon the group of a superior intelligence. The explanations given are credible, as given above.

The quality of historically preceding levels of intelligence have diminishingly different relational nexus. The further the superior intelligence is moved back in time on a vertical temporal axis, on any given plane, it seems as if superior intelligence has a positive, uplifting effect on groups. Superior intelligence before the printing presse was not generally repressed, as was the case with the Catholic Curch’s primal possession of knowledge, in fact populations preceding the dark and the Middle Ages had high regard for their poets, playwrights, philosophers. The separation between the individual and groups was pretty much demonstrated as based in the kind of mystery that Plato’s Meno could demonstrate. If a slave boy could gain a priori knowledge, in a society where the only difference to the acquisition to rights were the privilege of citizenship over slavery, then surely, one could conclude that intelligence qua the breeder of knowledge could be appraised as a positive thing among the underprivileged-uneducated-unintelligent.
Here the acquisition of intelligence did not yet evolve into the realm of the genetic theory, in fact quite the opposite.

Inherited traits followed simply along the same route as the Divine Right of Inheritence of political power, they were made simply to co-exist upon the backs of general theories, to supplant and justify them.

Therefore, it could be claimed, that the imminent level of argument may be acquired only from the perspective of another current psychologically narrow minded aspect, that is, that the democratic masses do not, can not tolerate the super intelligent, for very basic reasons quite rightly listed above: in ter,s of utility, and lessening functionality based on immature psychological defenses, such as vanity, envy, and the such.

Nice piece of philosophy btw.

/or, it is better to have smart individuals, because to get that requires a greater field of variance [or why not both?]. This equally yields a larger repetoire, a greater pool of creation if you will.

Unless you are an intellectual Pharaoh. I agree that if you said to everyone ‘you can all choose’ [a given thing], then you’d get a zillion different opinions. Try being the boss of that company eh! However, if you want that new product to stop the markets going sour [brains require continual refreshing of new information [like dennet said]], you need the creative opposite of singular dominion, or a pyramidic hierarchy et al. Most of our societies have both in some fashion, or they fail aka communism.

Individual vs collective is a far too simplistic view of the issue. Which type of collective organization and order is the issue. Distributed intelligence is essential even to the individual, but is only better if organized and ordered reasonably well.

And part of the issue of analyzing this situation is that by saying “individual”, one is presuming the lowest order, whatever that order might be, the “hu-man”. Nature would always have to have a lowest order regardless of how greatly intelligent any being became. A lowest order is necessary. So in the formation of Man, the human was all that was required, thus all nature did. Nature tends to be lazy.

People do the same thing when designing sophisticated systems. They choose a base level of intelligence (whether they knew it or not) and then start connecting them. Often the optimal level of intelligence distribution must be experimentally discovered so a cell phone might get too intelligent for its own good but through generations of challenges, the “system” ends up with the least distributed the most - hu-phone.

I would personally argue for a far better organization method than has been used and due to technology, that method will come about. The problem is that I don’t know that homosapian will be around to participate.