Sort of, depending on the model. For most models, if you take your "probability of dying in year X", and sum that over all the years from 0 to infinity, you'll get a 100% probability of dying at some point. The interesting thing is that there's no individual year with a 100% probability of dying - the certainty of death is just because "forever" is a really long time.
There are hypothetical distributions, though, where the sum total of probabilities is less than 100% - where some proportion of the population will, statistically, never die.
Of course this brings us to the real issue, which is that what the model says doesn't really matter - if the model disagrees with reality in extreme cases, reality wins.
There are hypothetical distributions, though, where the sum total of probabilities is less than 100% - where some proportion of the population will, statistically, never die.
Of course this brings us to the real issue, which is that what the model says doesn't really matter - if the model disagrees with reality in extreme cases, reality wins.