I think a better way to frame my argument is: look at the inflection points for each of the starting ages in the graph. 'at birth' has been going up since we ha

I think a better way to frame my argument is: look at the inflection points for each of the starting ages in the graph. 'at birth' has been going up since we have reliable data (1845). @20 has been going up since 1875. @60 has been going up since 1915. @75 has been going up since 1945. @85 only started to go up around 1980. @90/@95 look like they've only barely started to go up in the last decade. @100 and above are still completely flat. 100 years ago you could have looked at this data and just as easily said that there must be genetic/resource barriers to increasing life expectancy for 60-year-olds, since we've never been able to make a dent in their life expectancy -- at the time, the graph for @60 would still have been flat. Today that is obviously incorrect. The same thing could have been said 50 years ago for 80-year-olds, and again, that's now obviously incorrect. It's entirely possible that your hypothesis is valid, but if I had to guess I would say mine is more likely. Put another way, it's possible that 50 years from now the graph will still be flat for 100-year-olds, but I see no fundamental reason why that should be true, and the data certainly suggests to me that it won't be. A good way to test this might be the following: Based on the rest of this graph, I would guess that a decade from now, we'll see noticeable acceleration in the @100 line. If we don't, that might suggest that there is some fundamental barrier, or at least some anomalously hard problems there.