> Besides, how could terraforming Mars be easier than fixing our problems?

We don't need to terraform Mars to our incredibly fine tuned biology. What comes next will evolve beyond our limitations. Humans are a stepping stone. Just a blink of an eye in an infinite time scale.

What? There is no reason to think humans are "a stepping stone," there are plenty of dead-ends in evolution. There's certainly no reason to think that putting biological things on Mars will in any way cause them to evolve into Mars-robust creatures. Evolution is a slow, slow, slow random walk throughout which nearly all conditions have to be nearly-perfect.

Natural selection is a metaheuristic maximizing on the landscape of fitness topology, available energy resources, etc. It relies upon biological constraints, such as existing genes, organs, and biological functions available. It cannot make fantastic leaps quickly that would require significant backtracking.

I'm not suggesting putting biological organisms anywhere. I'm suggesting that at some point in the future an AI will be able to rapidly modify its own substrate independent of biology, metabolic inputs, and genetic inheritance. That it won't be subject to information loss due to death or other physical constraints. That it will much more quickly be able to build support systems for itself to harness energy and raw materials wherever it goes.

Unless we kill ourselves in the near future, these systems will likely emerge and will likely proliferate faster than us. They're not going to be stuck on earth with us. There are abundant energy resources and building materials in the solar system and greater galaxy beyond.