The stuff I read in the morning is very different from the stuff I read through the day. The morning is all about news - where I am playing catch up to what the world's editors think is important for me. I hardly look back at yesterdays news. In contrast, what I read through the day is largely intent/search driven, suggested by a friend or colleague, discovered through email etc. This content typically has longer shelf life, and is more valuable to me.

This is why I think, what I read as "News" should be largely driven by what I'm doing throughout the day. I'm working on Pugmarks.me [1] with this exact premise. Here are some design principles we're using:

1. What I read should be greatly influenced by my professional interests, as opposed to what is popular 2. My colleagues often have an overlap with my interests, and their experiences should help me discover great content 3. If have a set of sources that I trust, and this is continually evolving. A system should attempt to help me discover content from these sources (and potential new sources) before reaching out to the larger web 4. When I discover great content, I should be able to hold on to it, and rediscover it when I'm in a similar context again in the future

With these principles in action at Pugmarks.me, I've been able to significantly improve the quality of my reading.

[1]: http://pugmarks.me