Hacker News Reader

Comment by robotresearcher | Hacker News Reader

> not signs of "collective intelligence", as much as they are a manifestation of good algorithm design, honed by aeons of evolution.

What is intelligence, if not that?

View on HN

Intelligence is the wrong word in this case. These are complicated emergent behaviours that arise from extremely simple rules applied locally and iteratively by each individual. There's no planning, forethought, or central driving will at play here.

If this is intelligence, then so is bone growth, which follows a similarly simple algorithm, is also honed by evolution, and is also inspiring real-world applications, e.g.:

http://www.engineering.com/DesignSoftware/DesignSoftwareArti...

https://www.fastcodesign.com/1671102/a-3-d-printed-house-tha...

There's no planning, forethought, or central driving will at play here.

Do there have to be these things to show intelligence? Simple markets without central planning absolutely destroy central planning in terms of efficiency:

http://www.realussr.com/ussr/queues/

Maybe the intelligent, planning, ants were overtaken by the unintelligent ants :-)

The same could be said for neurons.

Probably best not to circumscribe 'intelligence' even if the ant example is as you suggest. What is planning, forethought or central driving? Braitenberg nailed this back in 1986 in his book Vehicles in which he describes simple toys showing 'increasing intricate if not always successful or civilized behaviour'.

See chapters (vehicles) on 'Foresight', 'Egotism' and 'Optimism'. https://en.wikipedia.org/wiki/Braitenberg_vehicle

Your immune system is a manifestation of good algorithm design, honed by aeons of evolution. It does not follow that it is intelligent.

That's true but not relevant.

Flip-flops are a manifestation of human design honed by millennia of design evolution. Tennis balls are a manifestation of human design honed by millennia of design evolution.

It does not follow that flip-flops are tennis balls.

But both statements above are still true, and their lack of equivalent properties is not an argument against their common history.

Though my post is consistent with your last paragraph, I see what you mean. My intent was to clarify that the common history is not an argument for the ants' behavior being considered intelligent, but you can rightly say that you never made that claim. Cortesi's larger point, that complex behavior can arise from a few simple and inflexible rules that do not deserve to be called intelligence, is not being contested.

Cortesi seemed to me to be saying something different: that ant behaviour was not intelligent, but merely nice algorithms honed by evolution. Which implies that intelligent is not nice algorithms honed by evolution. And since evolution is the only mechanism widely agreed to have produced intelligence so far (if we ignore divine intervention), I can't agree with that.

Reply on news.ycombinator.com