2 ms·
There seems to be some confusion between complexity, quantity of DNA, and intelligence as a measure of organism complexity. A larger quantity of DNA does not t
by fchollet 12y ago
There seems to be some confusion between complexity, quantity of DNA, and intelligence as a measure of organism complexity.
A larger quantity of DNA does not translate into a more "complex" organism, as amoebas prove. In fact, I would think it likely that organisms with a very large codebase have to be very simple ones, since the simpler an organism is the "safer" it is to add useless DNA code to it without breaking anything essential.
And consider intelligence: it is generally thought that the human brain is largely self-similar (the same structures can be observed throughout the cortex, repeated in layers and columns) and therefore very simple from an information-theoretic point of view. You could express what it does using very little information, or code. The closest we have to an artificial visual cortex, deep convolutional neural networks (shown to have a representational power equivalent to the primate visual cortex [1]), can be implemented in just a couple hundred lines of CUDA code. Just because a system is hard to understand does not mean it has to be complicated, or that it requires a lot of code (DNA or otherwise) to implement.
[1] http://arxiv.org/abs/1406.3284 http://arxiv.org/abs/1406.3284