The quote most commonly attributed to ice hockey great Wayne Gretzky is that he skated to where the puck was going to be next, not where it was at the time.
Google has clearly taken this message to heart, to judge from the advances it showed off on Monday in the areas of mobile and real-time search.
What made this all the more striking was the contrast with Microsoft, which last week tried to whip up attention for Bing with a rare event in Silicon Valley to boast its own search innovations. Microsoft’s interest seemed very much focussed on where the puck already is – though it still has some fresh ideas that suggest it should be able to put up a better fight in this latest round of the Search Wars.
Voice- and image-recognition figured high in Google’s presentation. That’s ironic, given how much Microsoft has been talking up the potential for non-text based methods of interaction with computers. Google chose action rather than talk – though it relied in some cases on tech demonstrations of services that are not yet available.
Most attention-grabbing was Google Goggles, which uses image-recognition to return search results. Snap a picture with an Android phone, and a set of relevant search results is returned to the handset. This only works with newer models, though – as with its recent navigation app, Google seems to have abandoned users stuck with Android version 1.5 or earlier.
The responsiveness depends a lot on the speed of the network, but Vic Gundotra, the VP of engineering in charge of mobile development, told me Google could return results in 3-7 seconds.
The potential applications of a service like this make the privacy row in Europe over StreetView seem like a sideshow. The technology is capable of facial recognition, so in theory it could be used to make anyone instantly recognisable. Mr Gundotra said Google would not use facial recognition in any form, at least until there had been a chance for a wider public debate over the issue – though he told me he thought there might be some applications where people would welcome being “recognised” by intelligent machines in public places, for instance as they walk into a store.
In the area of voice-recognition, Google showed off a new Japanese version of its voice-powered search (to add to the English and Chinese versions), and a translation service that can turn speech in one language into another almost instantly. The latter was a technology demonstration, but Google said it hoped to be able to make it available as a public service before the end of next year.
Compared with attention-grabbing demonstrations like these, Microsoft’s own focus with Bing sounds almost quaintly old-fashioned. The main aim, its search engineers say, is to get users an answer to their queries faster, not just present them with a list of websites to trawl through.
That has led it to try to extract information from various sources – the Web, structured databases like Wolfram Alpha, and submissions from “partners” like the Mayo Clinic (though, despite the talks with News Corp, it has played down the suggestion that it is looking for exclusive content, whether or not it is prepared to pay to get it).
Is Microsoft making the classic mistake of the “fast follower” – trying to emulate a rival’s current success, rather than trying to anticipate what comes next?
That is definitely a risk. But it’s also the case that Google’s core Web search product is not unassailable.
If you want a direct answer to a question, Google sometimes offers one in its search results page – the weather forecast for London, for instance, or the GDP of the US. But its main philosophy, as Marissa Mayer emphasised again on Monday, is still to hand users off to other Websites as quickly and relevantly as possible. In many cases that leads to Wikipedia – one of the Web’s most important resources, but also one of the most problematic.
Microsoft is right to sense there is room for something better. But when it comes to mobile, it will have to run much faster to keep up.
