As you can see from my letter to President Obama, I am starting my company to build a nextgen internet answer engine. Tomorrow, I am meeting Mayor Bloomberg‘s Startup Accounts Director in the Upper Manhattan Office. Then I am going to participate in an innovative ideas brainstorm meeting and then an internet startups networking meeting, both on East Houston Street in Manhattan. Here is the news I would like to present for you. This is our chance to recover from the current national and global mess and jump into a nextgen smart life.
Computer Scientist Calls for Natural Language Q&A Searching
Oren Etzioni, a computer scientist at the University of Washington, has made a call for deeper investment in actively providing answers in a natural language formats. This, claims Etzioni, is the future of search.
The “Natural Language Q&A” Concept
Etzioni’s article claims that the blue links found in search engines will soon be a thing of the past. Instead, he claims, we’ll be typing in natural-language questions and getting understandable answers back. He points to recent technological breakthroughs – including the Watson computer system that was capable of winning at “Jeopardy!” – as signs of things to come.
He states that the search engines currently provide the “electronic equivalent of the index at the back of a reference book,” when they could be connecting users directly with the responses they need. He’s not assuming that search will change overnight.
However, as an engineer who works on natural language technology, Etzioni feels that the search engines need to start investing more heavily into the idea of natural language responses. “We’re under-invested in this technology of the future,” he states.
From Information to Response
Etzioni isn’t alone in his opinion. Vivek Wadhwa, in his famed “What I Want in My New Google” article, talked about much the same thing. Bing, meanwhile, is focused on becoming a “decision engine” that goes beyond merely providing your standard 10 URL results.
There’s certainly a lot of technology surrounding automated comprehension of user questions. In some cases, Google, Bing, and others already give a direct response. For example, if you pop over to Google now and type “time in Milan, Italy,” Google will provide a widget at the top that gives you the precise time.
That’s easy for some things, but what about general information searches? I often type in the name of a business or individual to get an idea of where my hunt for full questions should start.
Additionally, answers are often diverse and contradicting. For example, try the search “Is Obama a good president?” and you’ll get plenty of answers. Good luck to a search engine trying to answer that one.