Personally, I read a lot, and write some code. I need to write more code than I read, but its a good start. And a lot of what I read does not have to be necessarily technoloy specific, rather books like "Clean Code" and "Code Complete", Intro To Algorithms etc, patterns, architectures etc.
I also tend to read a lot of self-help (GTD), business (Good to Great) and career specific books (Career Intensity) and similar blogs and sites.
In my experience, it helps to try and really really practice something. As Bruce Lee once said, "I fear not the man who has practiced 10,000 kicks once, but I fear the man who has practiced one kick 10,000 times."
To give you an example, I am currently reading Clean Code, and one of the pieces of advice is "A method should do one thing, and one thing only". Nothing new there, but normally I do take the time and effort to keep my code short and clean, but this book says that exception handling (in Java) IS ONE thing. So your method should not attempt to do exception handling as well as some other piece of functionality. For the past 3 days at work, I have tried to do that, not because the book says it, but because I did not think of it that way, and I want to give it a whirl before I make my mind up.
A lot of people I know don't read, and even those who do, don't practice it. If you want to grow - learn, practice, measure, then either accept it, or throw it away.
Update - Other HN readers have other pieces of advice that I did not add here. But working sample projects, reading documentation, tutorials online (I refer to them as the magazine version of tech books :D) all help
Really though, it's how I learn about what's going on. I don't necessarily learn every new framework that's released here or become an expert on every new web tool - but at least I'm exposed to it.
And I also do the exact opposite: take breaks away from all stimulation and just code on a project until it's done, preferably in something new (framework, language, paradigm) or at least in a new domain.
I also try to blog and write down my findings which helps me further understand everything especially when I get feedback.
This week I spent time getting dual monitors working on my Ubuntu partition. It was rough because my ATI card does not handle well on Linux so it took a lot of reading, trial and error, and understanding of how it all works. Now, I am writing this from Ubuntu and have my monitors working great.
It was essentially an assignment. I set a goal, worked on it, studied for it, and "finished" the assignment by getting it working and now I have all this new knowledge about xorg and ATI cards on Linux that I thought I would never care about
after reading 'ecology of planted aquarium by Diana Walstad', i basically say bye bye to weekly water change, lighting, aerator, pump, fertilizers and other fancy stuffs
now i only have soil (any, cheap will do), gravel, water, plants and fish - the fish food and mulm become fert ... along with semiannual water change
the algae problem is combatted with fast growing emergent (root+in-water+out-water stem) plant ... the plants will take nutrient faster (from soil+water+air) than the algae (from water alone) ever will
relevance to my work? now i use fewer, simpler (but powerful) tools with rapid iterations (hopefully at faster rate than bug creation)
and it worked wonderfully. i get to hack around with cutting edge stuff, building prototypes to see if it works for us.