Hi Brett,
Until Jacquie has a chance to respond, here's my tuppence worth (or $0.02).
There's another discussion going on about the relative merits of general skills vs. specialism, on this
thread:
https://coderanch.com/t/471396/Jobs-Discussion/careers/Wide-Shallow-VS-Narrow-Deep
"Rockstar"? Be good at what you do now, then build on that. The last thing this industry needs is more over-hyped "rock stars"!
As for the new skills, think about the underlying concepts and pick out the themes you want to explore in greater depth e.g. dynamic languages on the JVM. Then pick a corresponding tool to use for your practical investigations e.g. Groovy, JRuby, Scala, Jython etc. Buy a book or use online resources to get you started, pick a simple application to implement using these new toys, preferably something that might be relevant to your day-job, and go ahead and build something real in your spare time. That's the only way to learn how these things really work, and you may be able to use it to demonstrate to your employer why they might find these technologies useful.
And remember, your employer may be quite right to be sceptical about these new toys. As techies, we often get so enthusiastic about new technology that we forget the primary purpose of IT in most businesses is to support the business, not to provide a playground for our enthusiasms. If you can show them how their business might benefit e.g. from improved productivity in IT development, or more maintainable systems, perhaps on a smaller
test project, then they might be more interested. At the very least, they might appreciate you spending your own time on thinking about practical ideas to improve their business.
Also, most conferences and symposia are as much about marketing as anything else. There's a new language or tool every few weeks, and everybody's selling their own brand of snake oil. Start looking underneath the hype and try to spot potential practical applications e.g. in your own working environment, as well as the signs of any genuinely new concepts or paradigms (as Jacquie says elsewhere), that might be more generally applicable.
Good luck!