Peter and I had an ongoing argument over determinism - LaPlace's Demon was the creature that had a perfect knowledge of the past and present, so could fully predict the future. Peter thought we might eventually be like the demon, that our inability to predict merely reflected the imperfection of our knowledge. I never agreed with this - there is always an element of surprise in nature. Many natural systems are organized, but this doesn't make them machine-like. We can't predict the next speciation event. If we plant an acorn, we'll get an oak of a known species, but we can't predict the exact form. We can predict a hurricane, but its behavior may be partly unknowable.

Reductionism is a concept related to determinism, and many biologists assume that evolution and the resulting organisms can all be reduced to DNA. Early biologists thought perhaps tiny humans, homunculi, existed fully formed in sperm cells and directed development. More recently, the collection of molecules that is DNA has been given the role of the "ghost in the machine." This reductionism isn't helpful, but it has become very common, usually taking the form of Richard Dawkins' selfish genes and offering an inaccurate view of biology. The self-organizing and unpredictable aspects of nature are related to the idea of emergent properties where the whole can't be explained or even described exclusively in terms of its parts.