Google has announced its latest AI model, Gemini 1.0, which appears to be one of the most capable artificial intelligence models. As Google figures…
When Google launched its search engine a little more than a decade ago, the internet contained less than a billion individual pages.
Today, the web grows by several billion pages each day – pages that are ever-more complex, dynamic, and interactive. While organising this information grows more challenging, users’ expectations only rise: today people expect the right answer immediately, in the first result and in the language of their choice. People want to be told if there are things about their search that they failed to consider. Don’t just tell me what I think I need to know, tell me if you know something that I don’t know I need to know.
Although Google and other search engines have improved in sorting through this great and growing heap of information, we believe search is far from a solved problem. We must become better at providing answers to complex questions, and we have to anticipate our users’ needs. At Google we think we’re just scratching the surface of what search engines will be able to do in terms of providing customised, personalised, and relevant results.
We are not alone in the quest to organise the web and many other constructive approaches are appearing. Most recently, Microsoft launched it’s new search engine called Bing; and the renowned inventor Stephen Wolfram launched the computational engine Wolfram Alpha. These two engines take different approaches to the problem of search. Bing does new things with respect to categorising data and presentation of video, image and other content. Wolfram Alpha is innovative in that rather than trying to aggregate data, it attempts to analyse data, and provide a broad picture in answer to a given search query.
The user wins
These new entrants are welcome. We all know that competition breeds innovation, particularly when it comes to web search. Internet searchers are always looking for the search engine that will give them the best results. For instance, earlier this year Forrester research of US searchers showed that fifty-five percent use more than one search engine every week. This is great news for web surfers because it means that all companies involved are compelled to work harder at providing the best possible search experience and results. The web searcher wins, and the world’s information (at least that part of it which is digitised) becomes more accessible search-by-search.
The next great innovation in search can come from everywhere, and competition is a great motivator to work hard to be the company that comes up with that innovation. In the end, many different approaches to search probably will coexist. Unlike in the Highlander, there can be more than one, and there will be more than one.
To the person searching on Google, the back-end of what we do is likely of little interest. But the right answer can’t be found through one simple approach. For example, “Where can I read the latest news about Olympic gymnastics?” requires a very different kind of technology approach than “What’s the difference between the most popular models of hybrid cars?” or “What’s my nearest pizza restaurant that delivers?” The best answers to these might be news sites, graphs, tables of collated data from all over the internet, or maps. They might also be answered better based on your location, or previous searches you might have done.
One of the areas we think has a lot of promise is search personalisation. All the new approaches to search aim to some degree to figure out better what the user means when he types a given query and to be able to provide the most correct answer in each case. The perfect search engine will understand what you meant, not simply what you typed.
We’re working on various new tools at the moment to fulfill this aim. For instance we recently launched the search options feature, which allows you to track down information that’s hard to get at using a single search term. We also launched Google Squared, an experimental product that attempts to answer queries whose answers aren’t to be found in a single web page, but rather scattered all over the internet.
We continue to invest in promising new areas, and we’re looking forward to offering new innovations. User expectation increases in parallel with the capacity of search engines to answer new and difficult kinds of questions. In addition, finding the next great innovation means anticipating needs users’ never knew they had. Before the internet took its modern shape, could you have imagined being able to call to hand all the facts, images, videos, and other information as quickly as you could think of it?
We’ve seen larger and smaller leaps since the inception of the search engine, and we’re all attempting now, by large and small steps, to come up with the next great thing that will change and improve the way we access information.