> Don't use Google search.<p>This is easier said than done, as the other search engines are still not as good as Google, even though Google's results have been getting worse. This may be a controversial opinion, but it's not what my comment is about. I'm going to make a more scandalous suggestion.<p>Don't use search engines at all.<p>The idea that a centralized one-size-fits-all search engine is necessary is preposterous. The Web makes available all kinds of information, and unifying it all under a single data model is difficult, and doesn't even make sense. (Does anyone remember the semantic web?) Unifying the world's information behind a single search facade is likewise a Very Difficult Task (TM), one that's likely to fall into the trap of big business, as search has done, because the required resources are so huge.<p>But what if it's solving the wrong problem? Information of a particular type tends to gravitate to local centers of storage, so to speak, which are specific to the type of information being stored. For example:<p>- Encyclopedic knowledge is in Wikipedia.<p>- You can find places by searching Foursquare, Yelp, Apple Maps, OSM, ...<p>- Q&A about programming (and lots of other topics) is on StackExchange.<p>- News aggregators have been beaten to death, and multiple are available.<p>- You can search Twitter using Twitter, and Facebook using Facebook.<p>I can go on, but the point is clear: every single Web-connected system offers a search function of its own, one that's likely specialized to the type of information stored in that system. It'll most certainly do a better job at searching that local store, and will do so more quickly and cheaply than a centralized, generic search engine. This also avoids the moral hazard of search centralization.<p>This leaves the little guy: the random small website or blog, where the majority of true gems are found. Google locates these by sheer brute force: they literally index the entire web. They've taken a relative eternity to do so, but it's a problem that could have been solved by something better than mere force.<p>Does anyone remember webrings? <a href="https://en.wikipedia.org/wiki/Webring" rel="nofollow">https://en.wikipedia.org/wiki/Webring</a> What if "the little guys" organized in webrings and directories? This doesn't seem like a technical problem, as a webring or directory is trivial to build. Could this be a UX problem that hasn't been solved to the satisfaction of a modern Web user? Is anyone or anything taking another stab at this?<p>In closing, I'll throw out one last vague notion: that of an openly federated search. How cool would that be? We don't need Google for that at all.