[marvin]

The thing about "unknown unknowns" is that they are unknown. It's like asking "what could possibly be outside the edge of the map, surely it's more of the same".

This isn't a matter of water flowing downhill through a pipe or a piston moving through a cylinder. Although we have a good understanding of the process, we don't have decades of extensive engineering experience to give us empiric reassurance that there are no gotchas. We didn't put a billion people into airplanes nine months after the first commercial airliner was developed.

Sure, hopefully our current understanding is good enough and there are no long-term side effects. But it's naïve to assume a likelihood of 100%.

Although the elderly and sick have a high likelihood of a serious covid case, young and healthy people are exceedingly likely to have a mild disease. These probabilities are what you have to judge the confidence of long-term vaccine safety against.