I liked Von Neumann's analysis of the problem as relating to the size of the effects available to us. As technology progresses, the effect size increases, as t

I liked Von Neumann's analysis of the problem as relating to the size of the effects available to us. As technology progresses, the effect size increases, as the effect size increases, the danger of mistakes gets larger. Unfortunately "The most hopeful answer is that the human species has been subjected to similar tests before and seems to have a congenital ability to come through, after varying amounts of trouble" is not actually a hopeful answer (although it might have been the most hopeful answer available to him at the time). It's not hopeful because any amount of trouble can be catastrophic when the effect size is larger than the size of the ecosystem that supports survival. The only approach that seems to me to have a chance of a modicum of success in the long run is to rapidly expand the size of the survival ecosystem. We need to develop the technology and political will to send out multiple, significant sized, independent generation ships or colonies with such a variety of governance systems and cultures that dangerous memes would be unlikely to spread easily between them. Physical and information communication between them and back to Earth would need to be infrequent and treated carefully, but with appropriate safeguards, the risk/reward would probably still work out that they could help each other and Earth. I don't believe any other approach to existential risk is likely to work. At best we'll resolve the existential risks relating to our current technology, but effect sizes won't stop with those, they'll continue increasing, and to survive long term, humanity needs outposts continually beyond that increasing bubble of effect size.