Chemistry guides evolution, claims theory: "As these primitive cells, or prokaryotes, extracted hydrogen from water they released oxygen, making the environment more oxidising. [...] But these oxidising elements could also damage the reducing chemistry in the cytoplasm. [...] For protection, there was just one option: isolate the elements within internal compartments, says Williams. And that gave rise to eukaryotes - single-celled organisms with a nucleus and other organelles". And then later, "Williams admits their theory has limitations. For instance, he agrees that Dawkins's argument is correct in that chance events drive the development of species. But he does not believe random events drive evolution overall. 'Whatever life throws away will become the thing that forces the next step in its development'."

So two interesting things in this article: first is that the choices made in one step of the evolution of a system become embedded in the environment for the next step (I'm happy with that). Second is the claim that this somehow removes chance, because if the system can only respond to the environment in a certain way, and the environment itself is actually a product of the system over time, then the path of development is inevitable. Which I'm not happy with. The mechanism of change isn't push, it isn't deciding on a solution and working towards it: it's pull, making all possible changes at once (given a large enough pool), and only the fit-for-purpose surviving. It's typical of what bugs me about New Scientist. A sensible, insightful idea (breaking the system/environment divide) spoiled by sensationalist journalism (the end of evolution!!!).