The technology does not change as fast as it claims to. There are new languages, frameworks and even concepts all the time, but most of them are the old things with a twist and most of them do not even catch on. So if you don't want to be living on the bleeding edge, there is a good strategy: stick to the stuff that has been time proven. An indicator that something is good and works is that it's considered old and uncool. Keep an eye on technologies that are being widely used and when they become practical standard (there should be lots of documentation, tutorials and books around this time) - learn if you could use it and if so then learn the technology.