I've actually tried this before, under the notion that I'd get a higher chance of landing the job relative to other applicants. I wouldn't do it again - I'm pretty sure the real outcome is that I lowered my perceived value. Signalling!
Now that I phrase it in light of game theory though - if it had worked, it would be a neat demonstration of how two super-rational players win out over game-theoretically rational players. What I describe is also how the free market is supposed to work - my benefit in "cooperating" derives from cutting the competition out of the trade by offering a better deal.
It looks like there's no incentive for them to post an honest salary range, and there's no incentive for you, having been told the salary range, to be honest about what you think you are worth.
It's not defecting, just making a choice that is strictly nonbeneficial for the other party.
If funding were available, the Centre for Effective Altruism would consider hiring someone to work closely with Prof Nick Bostrom to provide anything and everything he needs to be more productive. Bostrom is obviously the Director of the Future of Humanity Institute at Oxford University, and author of Superintelligence, the best guide yet to the possible risks posed by artificial intelligence.
Nobody has yet confirmed they will fund this role, but we are nevertheless interested in getting expressions of interest from suitable candidates.
The list of required characteristics is hefty, and the position would be a challenging one:
The research Bostrom can do is unique; to my knowledge we don't have anyone who has made such significant strides clarifying the biggest risks facing humanity as a whole. As a result, helping increase Bostrom's output by say, 20%, would be a major contribution. This person's work would also help the rest of the Future of Humanity Institute run smoothly.