I was reading a HN page http://news.ycombinator.com/item?id=898675 where Linus argues that Linux was not "designed" and was struck by the strength of his views. Here are some quotes:<p>"I'm deadly serious: we humans have _never_ been able to replicate something more complicated than what we ourselves are, yet natural selection did it without even thinking.<p>Don't underestimate the power of survival of the fittest."<p>This intrigued me so I looked at his blog http://torvalds-family.blogspot.com where he comments a few times on what he is reading. I will list some of the books he mentions there:<p>Phantoms in the Brain,
The Brain that Changes Itself,
Why Evolution is True,
Genesis: The Scientific Quest for Life's Origins<p>This quote is interesting:<p>"Usually, I tend to read about genetics or similar (that is, when I read anything serious to begin with, which tends to be less than 10% of the time). This one is obviously related, but about the processes that came before it all began. And it also gives more of a look into the issues faced by <BOLD> somebody </BOLD> trying to do experiments in the area."<p>From the above comments his viewpoint possibly falls into the artificial life camp that tries to imitate traditional biology by recreating biological phenomena (http://en.wikipedia.org/wiki/Artificial_life).<p>What do you think of this idea? He is well-noted for his programming skills and general ability. Maybe the kernel is just a basis for some sort of massive ai-life simulation?
There is no artificial intelligence, just a lot of humans doing experiments and keeping the best results in the kernel. I think that's plain-old "intelligence".
<i>Maybe the kernel is just a basis for some sort of massive ai-life simulation?</i><p>The kernel isn't the emergent life, the community is. The kernel is just the/(a) side effect.
He said he believes in guided evolution, which I agree is the best way to do things. But really it's the only way, because if you're not doing that then at some point you must have lied to yourself when your vision misaligned with results.<p>I think what he's going at is trying to design things logically and rationally, just as evolution does. That is, seeing how the world designs itself and modeling his own methods to it with a little human guidance to speed the what would be random process up.
sounds like he's also been reading wolfram's nks. the whole book is about how complexity is created. and wolfram suggests that humans wouldn't be able to <i>purposefully</i> create something more complicated than ourselves (though leaving open the possibility of doing it by semi-accident).