Google Search Future Channels 'Star Trek' Computer
Google's top search engineers, including Google Fellows Amit Singhal and Ben Gomes, run through a video snapshot of the company's search engine evolution over the last 13 years.
None of it is secret, though some of it may seem to have been resurrected from obscurity. Check it out for yourself:
Where search came from doesn't interest me. Search as it is today doesn't thrill me. It's a tool, gets the job done. Where search is going promises to be more interesting.
Singhal, who noted users require more complex answers to longer queries, said he dreams of building a "Star Trek" computer.
In my ideal world, I would be able to walk up to a computer and say 'Hey, what is the best time for me to sow seeds in India given that monsoon was early this year.' Once we can answer that question, which we don't today, people will be looking for answers to more complex questions..."
If you grok that search today requires hundreds of signals for accuracy and relevance, imagine what the future of search will require. Thousands, if not tens of thousands of signals.
The query Singhal envisions being able to ask would have to take into account date, time, Farmer's Almanac type weather stats and other variables.
Google Voice Search has the potential to alter the future of search input. In its current instantiation, Voice Search is pretty hit and miss.
There are accuracy and noise interference issues. Microphones and software haven't evolved enough to communicate spoken words properly. For example, if everyone is speaking voice commands all of the time, it could get noisy.
At some point, we may simply speak all of our searches into our computers, smartphones, tablets and -- I'm reaching a lot -- refrigerators, lights and other home appliances. This will save us a boatload of time.
Voice recognition -- recognition of who is speaking, not just what is being spoken -- will also become paramount. So will greater, artifical intelligence.
Can't get enough of the ins and outs of Google search? Read Singhal's post from last week on personalized results.