our example v. probability and the Fermi Paradox
We (or possibly, our AI descendants) are about 100-200 years away from sufficient solar system colonization to survive an earth-destroying Big Mistake. Over the 10k years of technological development, that's pretty close.
It's no guarantee that we'll make it, but if there are more than a few other alien civilizations out there, then probability suggests someone's going to luck their way through the bottleneck of sufficient technology to destroy the home planet while not yet moving off-planet (I'm assuming that having witnessed the mistake done on their home planet, their error-repetition rate would slow substantially). In other words, our example doesn't suggest that a planet-destroying technological mistake explains the Fermi Paradox.