As you know, Google announced their Hummingbird algorithm about a month after it launched, claiming no one noticed and no one should notice. But we do think we did notice but no one can confirm that outside of Google and they won't.
Some forward thinking SEOs and webmasters are already thinking up what the end game for Google is with Hummingbird and how to adapt their sites to fit that box.
A WebmasterWorld thread has some really interesting conversation around what some believe the key difference is before and after Hummingbird.
Unique Content versus Useful Content
While unique content is more of a Google Panda related thing, useful content although Panda, is maybe more Hummingbird.
Google understands searchers queries differently with Hummingbird than they did before. So how can the search results not change. How can you as a webmaster change your content to make it more useful, while it still being unique, to encourage Google to show your site over your competitors.
WebmasterWorld moderator, Robert Charlton, explained it so incredibly well:
It's no longer just a single page and its title satisfying a query... It becomes a whole site satisfying a range of users. With that kind of scope, the individual referrers are both less easy to specify and less determined by the landing page itself. Actually, not so different from what some of us have been preaching.
Don't optimize for keywords, optimize for a satisfied customer from stage one of the buying cycle to the end. Is it that easy? What if you don't offer all the stages? Well, I assume that is not exactly the point.
Robert believes this will eventually lead to search results that are "less a collection of content farms and more a collection of pages created with the user genuinely in mind." I am not 100% confident.
Keep in mind, this is just one theory of many and for the most part, the search results did not change that much compared to let's say Penguin 2.1.
Forum discussion at WebmasterWorld.