That seems more to be an assumption you're bringing in, rather than others. There's a huge gulf between "technology that can read the structure of the brain and recreate it" (which seems to be something merely lacking in know-how and tech, but with no fundamentally-impossible steps), and "simulate the entire universe as-is, in situ, and assume it's 100% accurate, and pluck people out of it" (a la https://qntm.org/responsibility), which seems outright impossible as no system can simulate itself fully. I'm honestly having trouble wrapping my head around how you're both managing to insist on the latter being the only possibility while simultaneously telling people that it's a bad assumption.