Is a Boltzmann brain by definition unstable? Over what timescales? While it exists, does it function within its presumably more stable context, or is it somehow

Is a Boltzmann brain by definition unstable? Over what timescales? While it exists, does it function within its presumably more stable context, or is it somehow magically disengaged from its context in a way that necessitates instability? What I'm getting at is that the idea of a Boltzmann brain, as juxtaposed with a "real brain", seems a little too " classical binary logic" for a universe that evolves. Perhaps some of our most resilient features have evolved with the help of spontaneous Boltzmann-brain-like intuitions, whims, or setups, but have been since then integrated in a stable way. I imagine a Boltzmann brain to be something like a spark, but not immune from outside interaction or influence. Am I missing something? What happens if a Boltzmann brain "catches on"? Is it still a Boltzmann brain, now that it has contextualized itself in a stable way? Was it ever a Boltzmann brain? This also comes down to the meaning of "randomness", and I don't think the mathematical jury is quite out on that subject.