[api]

My favorite version of this classifies life as an exotic phase of matter in the sense of other exotic phases like Bose-Einstein condensates, etc.

Life is a phase of matter in which the dynamics of information and Turing-complete computation overcome the ordinary dynamics of matter and energy. I am not aware of who originated this idea, but I heard it from Christoph Adami (a researcher in evolutionary information theory). I also recall someone at a conference proposing the term "Turium," for "Turing-complete phase of matter."

Earth just happens to have an environment that contains a large abundance of this phase of matter, much like how stars are full of plasma etc.

[kubanczyk]

How is this model useful? It certainly moves away from our domain knowledge and paints a new picture that is considerably more fuzzy than the old one. What do you gain by considering life in such way?

Side note: computation seems like a wrong term, because it's not about counting/calculating/determining any abstract value, is it? A Turing-complete mechanism maybe?

[api]

I find the idea of it as a phase of matter very conceptually elegant in that it neatly classifies life as a unique but still physical phenomenon. Nevertheless I agree with you that the details of the definition are hard to pin down.

Of course I have yet to see an attempted definition of life that isn’t a little squirrely or that isn’t just a list of criteria taken from the one example we have.