Obviously, there's the huge advantage of having the very latest version number - nobody can say your system's out of date!
I make it a point not to upgrade software unless I know of a particular benefit I want or that may have some value in the future that's worth today's pain. Databases in particular are things you should never update on a whim as usually you're using a database for a vital reason and something could change in an upgrade that upsets the apple cart.
UPDATE:
Notice that I didn't edit or delete the post just because three of you decided to vote this one down. That's because those who voted this down MISSED THE POINT and they might very well learn something by stopping and thinking...
The very idea of always going to a younger release is flawed; vendors love it, but if there's nothing wrong, don't fix it! This is especially true in the database world where it's not at all uncommon for entire enterprises depend on a database. Upgrading without knowing exactly what benefits and risks there are is just plain stupid.
The original poster said they didn't know if there were any benefits to be had. In my opinion it's VERY risky to suggest to that person that they upgrade because the risks are unknown. Sites often have dependencies they don't even know about! ...Hang out in this industry for some thirty years and maybe you'll figure this out... I guess there's also a big difference between coding engineers and the support staff that have to keep systems running 24/7. Those of us who have to keep the lights burning know how vital our systems are and that we have to take a more cautious approach than someone writing code.
Another point here; when you see something you don't directly agree with, you might just think about it for a while instead of just deciding it doesn't fit your paradigm so it must be wrong - and therefore voting it down. Voting down should only be done for things that are clearly wrong and not because it disagrees with your OPINION.