I would disagree. Sure the app itself does something ridiculously trivial. But setting up an app that is expected to scale is not (I assume it is, if they got 1.2 million). They must have some very good marketing, or something special to get that million.
I do the database for a DNA sequencing centre. I guess some of these things are considered "important" problems (working towards curing cancer, or other diseases). But most of the bioinformaticians are second rate programmers (there are a couple of geniuses thrown into the mix as well). Their work generally helps, but to say they are doing something smarter that a scalable web app is a bit naive. Most of the stuff is 100 line scripts that count some numbers, and do some basic stats. Their software doesn't scale, it doesn't even have to be reliable, as often it will only be run on a couple of datasets. At the end of the day its a different set of knowledge and problems, the only similarity being that there is some programming involved.
Ok, I didn't claim that starting Yo was easy, or that merely contributing to important problems is harder than starting Yo. I meant that, e.g., solving the energy problem is much harder than starting Yo. Basically, those big problems are so big that even if everyone at Yo devoted their careers to one of those problems, the chances of that being the difference in solving it is almost infinitesimal.
There have been a lot of brilliant efforts at curing cancer. You could spend a large portion of a brilliant career following up on one lead. Odds are, that lead won't yield a cure. Cancer is that hard.
I do the database for a DNA sequencing centre. I guess some of these things are considered "important" problems (working towards curing cancer, or other diseases). But most of the bioinformaticians are second rate programmers (there are a couple of geniuses thrown into the mix as well). Their work generally helps, but to say they are doing something smarter that a scalable web app is a bit naive. Most of the stuff is 100 line scripts that count some numbers, and do some basic stats. Their software doesn't scale, it doesn't even have to be reliable, as often it will only be run on a couple of datasets. At the end of the day its a different set of knowledge and problems, the only similarity being that there is some programming involved.