This is my 2nd year in the lab with 1 year remaining in the contract. The PI is a nice guy and well-funded, but he has completely shifted his focus to his business and clinic trials. For example, I just waited a month for him to start revising my manuscript. I'm unable to get any instructions or ideas from him. Now I've finished the only paper and have no promising project in hand, I want to quit and find a new postdoc. But I don't know how to ask my PI for a reference letter. What should I do to avoid burning the bridge? I think he even expects me to extend the contract for another 2 years. I don't have any job offer yet, but I feel like I should tell my PI before sending out resumes.
Tell him you want scientific variety, geographic variety, etc. I wouldn't get into the "he doesn't have time for you". Just indicate you are interested in new things. Tell him this is a natural break point since you wrapped one thing up, so it's not even like there is a huge dislocation. Tell him you plan to continue working on current contract but looking for other things, very actively, at same time.
[Keep soaking up the current salary. Do some things to keep him/you happy, but it doesn't sound like there is huge supervision, expectations, anyways. Assuming there is no new major project, figure out how to dust off some minor aspects and get some "cheaper" papers done that pad out your resume, build off of last project. Helps you out and the advisor at least sees some pub count even though you aren't solving more Millennium Prize problems. If he gives you a new big project, I would be OK doing that, but keep looking for next spot and be willing to ditch it if you need to hop to another assignment. Be able to write up incomplete work...there is an art to this and it is not as sleazy as it sounds. In fact a lot of work is wasted because people don't write up incomplete work and it sits for years instead or gets redone.]
Ask for his assistance. He'll either give it or won't give it. I suspect he will at least give his blessing. In any case, once you have told him, you just execute your plan and start looking for things. 2:1 anything you come up with ends up needing some time to develop and there is a natural dovetail with current contract end and new contract start.
Personally, I think a postdoc (or senior grad student) should be relatively independent and capable of starting, executing projects on own. And of writing papers that just need a rubber stamp from the advisor who attaches his name to pad his CV. (Definitely work on your paper writing...publications are how people keep score in academia and a PI is expected to be able to write a paper without someone above them.)