[t8sr]

I don't think I expressed myself very clearly, but I actually agree with you on most counts. The definition you listed is also the one I mentioned + evolution.

> What about fire?

I actually rambled about fire, and then deleted the paragraph because the comment was already all over the place. Fire, to me, is a good example of the "good definition" problem I mentioned. I guess we'd all agree it doesn't respond to stimuli, and a necessary property of having a metabolism would be some kind of homeostasis. It also doesn't undergo evolution.

But rejecting it on those grounds doesn't feel complete. Let's say we find some chemical reaction that does self-regulate. This exists in a lab or computer modeling (e.g. https://doi.org/10.1073/pnas.2022987118). Let's say its properties are determined by some initial chemical markers: a basis for heredity. Would we now call it alive?

I don't think we would. I think our definition of life is "I know it when I see it" and the more formal definitions we've tried on in astrobiology are basically just retrofitted to the examples we know or can think of.