Find a niche that rewards experience in something other than code. This could be non-technical (e.g. "business"), or it could be something more deeply technical (e.g. scaling, machine learning, search, etc.) To some extent, the older you get, the more this happens, naturally. But you can and should make strategic decisions about where you choose to devote your attention.

The important principle is that you should devote your learning time to things that are more likely to survive. A good way to do this is to pick something that has been around for a long time. Knowing how to write an optimized linux kernel driver is far more long-term-valuable than, say, a Javascript framework. Knowing how to do quantitative marketing is even more valuable than that.

The absolute worst thing you can do is to chase the new shiny every year. If you do that, you will never be more experienced than the most junior member of your team.