It is optimizing survivability, which is a moving target as it depends on geography, ecosystem etc, so yes it's not running an optimization algorithm in a traditional sense since the function is changing as the algorithm runs. But locally (i.e. inside a geological era) species can be seen as solving certain physical/chemical problems e.g. locomotion, poison, electrocution, flight etc.
I think our narrative surrounding survivability optimizes for some types of things (things that have coherent identity over time). eg. "a snapshot of a tree in time" lives less long than "a tree" (many snapshots-in-time together) And I think certain problem descriptions and solitions are better fitted to such survivability descriptions. But I'm not sure if there is an independent thing than the observer-language complex optimizing for anything.
Like I feel like there is an inherent observer bias in the way survival is being defined, it feels more like "observable." Like in the local time of a tree snapshot (from its perspective), time may feel different than from ours observing it.
There is a single common ancestor of all living things today. This organism survived objectively better than any other organism of its time (unless it's the very first organism ever, which is very unlikely). Same thing goes for e.g. the last common ancestor of mammals vs an animal living this time that doesn't have an extant grandchild. I think it's possible to define survivability objectively using a little graph theory.
"There is a single common ancestor of all living things today."
No, there isn't. There's an ancestor population.
"This organism survived objectively better than any other organism of its time"
No, it did not survive. Being an ancestor is not survival. And this has nothing to do with "optimization" ... if evolution optimizes survival, then your "single common ancestor" predated optimization. You're just throwing concepts together to support a prior belief, but the facts don't work that way.
"I think it's possible to define survivability objectively using a little graph theory."
Maybe you should actually try it. Perhaps then you would notice that your notion of "optimized survivability" is about the future, but you're talking about graphs of the past.
> There is a single common ancestor of all living things today.
Are you sure about that? Viruses too? (depends on your definition of "living" of course) To clarify: I don't know if there is or isn't, I just don't think anyone does.
Yes we definitely do (not counting virus) because all living things share ~5/10% of their genetic material (like you share 10% of your genes with some bacteria) [1] which is very unlikely if it were purely random. This is of course one evidence, there are more evidence towards this. Read: https://en.wikipedia.org/wiki/Tree_of_life_(biology)
This is an erroneous way to view evolution. It's tautological that those traits that increase fitness (more viable offspring) become more widespread in a population. But those that don't disappear ... their survival is not "optimized". And all species eventually become extinct.
