If you re an expert in some field in your team (the DB guy, the UI gal, the search Guru.. you know), it also means you're a bottleneck, and people have to constantly wait for you to make decisions on important things.

Here are three reasons why you should stop solving other people's problems, and start pairing with them when they come to you with questions:

1. You will be reducing the risk factor in the project. The risk is that you won't be there when you're needed and people will be stuck without you (also called the bus factor). Can the team handle you being a way for a couple weeks? if not, that is a huge risk.
2. You will teach other people to solve their own problems, thus making them better at decision making. This will improve your status from "expert at X" to "Expert factory on X". You will have  force multiplier of sorts. Think world of warcraft and you getting one level up. You can start influencing more people and more decisions, by teaching them how to think like you. Now you're the person who you can bring into a team to make them better at doing X, instead of just solving X for the team. That might also mean a salary raise for you.
3. You will have more free time to do the things you really want to do. Instead of solving the same boring problems for everyone, you now have time to focus on more interesting problems, research or dive into new areas. Delegation is free time. Now you get to become better too!

So what should you do?

• Someone comes to you with a problem: "hey the build is failing and I don't know why. please fix it for me".
• You: OK, come sit with me and we'll check this out together. I want you to try and do this with me so you won't have to wait for me next time this happens, or so that I won't be the only person here who knows how to solve it"
• Them: "ugh, I don't have time for this now"
• You: OK, so come to me later when you have time and we'll do this together. I have something I need to finish anyway.
• OR you again: "So what if I wasn't here? that's a huge risk we can't afford. Let me teach you at least how to think about this problem and some common ideas to solve it. it will take five minutes"

Usually that breaks the ice enough. If not, keep at it for next time, but be PRO ACTIVE. No one will do it for you.