I agree this is the main potential failure mode of that plan. That said, the upshot of challenge trials is that we can get good data with very few people, so quite a bit less than 200 could suffice; I haven’t run the numbers. More generally, it seems like the sort of thing where a bunch of folks on LessWrong would discuss it for a while and figure out a way to do it which gives maximum information for minimum risk.
Sure. Only in that specific scenario, DIY vaccine, for maybe a small community, very safe, scales well in that range; i would say the value of the information you could get from a challenge trial seems low.
Instead i would just give it to everyone.
I agree this is the main potential failure mode of that plan. That said, the upshot of challenge trials is that we can get good data with very few people, so quite a bit less than 200 could suffice; I haven’t run the numbers. More generally, it seems like the sort of thing where a bunch of folks on LessWrong would discuss it for a while and figure out a way to do it which gives maximum information for minimum risk.
Sure. Only in that specific scenario, DIY vaccine, for maybe a small community, very safe, scales well in that range; i would say the value of the information you could get from a challenge trial seems low. Instead i would just give it to everyone.