Posted by: Prospector PJ30 APR 2014
A paradox is a “seemingly absurd or contradictory statement or proposition which, when investigated, may prove to be well founded or true”. For example, Peto’s paradox, named after Oxford professor Richard Peto, states that an expected correlation between animal size and cancer prevalence does not exist.
Because every cell in the body has the potential to become cancerous, one would expect larger animals to have a greater chance of cancer than smaller ones. But not so. Humans and beluga whales share a similar risk, while certain breeds of mice have a much higher risk. Some biologists believe that the lack of correlation in Peto’s paradox comes from tumour-suppressing mechanisms in larger animals.
The cells of more complex organisms generally contain more DNA, as one would expect, but there are many exceptions. The genome of the unicellular Polychaos dubium, for example, reportedly contains over 200 times more DNA than humans’.
In eukaryotes, much of the DNA apparently encodes nothing. But the quantity of this “junk” DNA varies hugely between organisms. Over 98 per cent of the human genome is non-coding DNA, for example, while only about 2 per cent of a typical bacterial genome is junk. This widely differing amount of unused DNA is one reason for the lack of correlation known as the C-value paradox.
Moravec’s paradox centres around a difference between natural and artificial intelligence, which means that high-level reasoning requires little computer power, but low-level sensorimotor skills need huge computational resources. Therefore it is easy for computers to process logical problems, such as devising chess strategies, but much harder to program a computer to walk or accurately interpret speech.
Hans Moravec, a robotics research scientist, explained this observation in a theory based on evolution. All human skills use machinery designed by natural selection, so that basic skills have improved over a longer period. Because abstract thought only developed relatively recently, its implementation should not be expected to be markedly efficient.