It really is not that important. Solve the atmosphere problem, and the surface radiation levels would be plenty low enough. Additionally, even without a magnetic field, it’d take on the order of 100 million years for Mars to lose an atmosphere if it were terraformed. Earth, due to Sun increasing in temperature, may become uninhabitable in about ~250-500 million years without active measures as well.

Humans will be long extinct in 500 million years. That's not really something we need be planning for.

But our ancestors need not be.

Also, that's something of a self-fulfilling prophecy. In order to be a long-lived civilization, species, or even biosphere, we must learn to think and plan far beyond one's own lifespan.

Anyway, you're proving my original point about the (lack of a) "need" for a magnetic field. It isn't necessary (on any kind of reasonable timescale), provided you have sufficient atmosphere. On Earth as well, the atmosphere is much more important than the magnetic field for shielding us from space radiation.

EDIT:Which isn't to say we couldn't build one. In fact, building a magnetic field for Mars (say, by many parallel superconducting cables around the equator) may be easier than making enough atmospheric pressure.

Creating an atmosphere and magnetic field for Mars should be easily doable, although I don't think either is necessary for human life on Mars, by comparison to ensuring our species survives for 500 million years. Only a handful of simple species have survived that long, e.g. sponges and jellyfish. I don't see why you think humans could do the same. But assuming we are at most .04% into our species lifetime seems ridiculous.

Feel free to calculate the probability that a randomly selected species will survive 2500 times longer than they have to present.

Why would we randomly select the species? We have agency, here... We can foresee hundreds of millions of years into the future in ways that literally no other species can (as far as we can tell, we're the only species that has developed the mathematical abstractions to even contemplate such numbers).

Anyway, the point isn't humans as a species, unchanging and eternal. But humans as the progenitors of future intelligent lifeforms that may indeed be around for 500 million years or whatever. Stasis isn't a requirement or even desired.

...by CHOOSING to contemplate survival on such timescales and making decisions based on that contemplation, we are, in fact, changing the odds of surviving that long. So a random selection is not appropriate...

You are assuming that abilities potentially unique to humans make us, or our progeny, more likely to survive long-term than a randomly selected species. I don't think that is well supported.

Also, why do you think that we have agency? That seems another unsupported assumption.

I could see a general AI, if one is ever developed, being more likely to survive that long, given we even consider it alive to begin with. But a hominin descendant seems incredibly unlikely.

I guess I don't see why you think contemplating long-term survival would be correlated with longer survival rates.

AI would count as a descendant of humanity.

But yeah, we have agency on this question to the same extent we or anything else has agency.