> A real concern is about safety of those vaccines as typically they require ~3 years of testing and a single fatality aborts their use.
Drug development is usually serialized because it's so expensive. Why produce a bunch of drugs if it's going to be thrown away? Operation warp speed said let's just throw money at the problem so we can parallelize as much as possible, knowing that if it doesn't work out (effective, safety) then we as a country just ate the cost of trying. My understanding is the clinical trial parts weren't any different than would normally be done.
AFAIK vaccine safety is usually determinable within a few months. That is to say, if anything bad happens, it'll happen soon and not years later. Because enough months had past with larger and larger groups looking ok, plus extreme effectiveness, then emergency authorization was passed.
You could say emergency authorization is alpha/beta. But that alpha/beta has now been rolled out to hundreds of million of people and the safety profiles haven't changed from the clinical trial modeling. And if the failure probability is all up front, and we've already passed those milestones, then are we really still in alpha/beta territory?
mRNA vaccines are also not a new technology. They've been used for years now.
> Would you give alpha version to people you love?
If the option is an alpha which is passing all the 'unit tests' so to speak, or let them get covid and have far higher risk of dying or permanent disability, I'd for sure choose the alpha.