The perception among lay users of the Internet is that Google’s omniscience is unparalleled as far as looking up information on the Internet goes. But with Artificial Intelligence fast evolving, inference search engines like True Knowledge and Wolfram Alpha have begun to throw up results that promise to enlighten the user faster and more intuitively.
Google primarily indexes websites and stores millions of keywords in its database, based on which it throws up results. For instance, if a user were to type in ‘Bombay’, Google would return all webpages with the term ‘Bombay’ in them irrespective of what context it was needed in. The information is all pre-fed into the database.
On the contrary...
Inference search engines are programmed to recognise words in a context. They, too, recognise only keywords from the input, but show direct answers to queries rather than direct you to webpages on the Internet that will.
So if a user were to type in ‘Bombay’, such an engine will recognise Bombay as a metropolis and will accordingly display all of the city’s attributes much as it is pre-programmed to do. Wolfram Alpha comprehensively lists out the population, current temperature, local time and approximate elevation among other things. Try entering a complex mathematical equation involving calculus or Sin and Cos values and the website returns three alternative representations, three integral representations and three of what it calls multiple-argument formulae. Simply put, direct answers. Such search engines help keep malware away, as they minimise your exposure to other websites.
Too much information?
In a Google search the kind of answer you get depends on the keywords you type in. And while you may be flooded with websites, it’s important to remember, excess information isn’t necessarily useless. You invariably come across a piece of information browsing through website that you never intended to directly search for, but found useful nonetheless.
If you’re not too clear on what you want, Google helps you navigate through even the most ambiguous searches and get to the answer. Wolfram Alpha isn’t too helpful as it throws up only a select few details.
According to Shreyas Srinivasan, a post graduate in Computer studies, “ Inference search engines are experimental AI implementations on par with maybe an Alpha or a Beta test of a new kind of software that use intelligent search techniques.
The ultimate aim of projects like Wolfram Alpha is to demonstrate a concept so as to gain the sponsorship of search engines like Google so the technology can be incorporated into existing search engines.”
So is Google still God?
According to journalist Aashray Hariharan, “Inference websites aren’t as informative and I only get specific information on them. I don’t even consider them as competition to Google.”
Shreyas, who has to solve mathematical equations in his work, says, “While engines like Wolfram Alpha and True Knowledge are based on some solid research, and are fun to work with because of their responses, they’re not so much alternatives to Google as they are the future of Google.”
The future of search engines
Wolfram Alpha currently performs less than one-fifth of the operations a search engine with its kind of technology ideally should. So it doesn’t really do to compare Google and Wolfram Alpha. The whole point of having inference search engines is to streamline searches, and make them as direct and intuitive for the user. The only thing they’ve still got to figure is how to give the user what he wants, without telling him that’s all he needs.