5 Things & the Future of Search
Craig Neville-Manning (Chief Engineer, Google NY office) spoke yesterday here at Idea Festival. He outlined 5 core things he wanted us to know:
1. Think broadly
"Computer science is not just about computers."
Computer science is about building things that change the world. It is about research. And, it is about people.
2. Enable others
Google Maps is an example. Technology’s value can be increased by allowing others to build upon what you have. They will build things you may not even have predicted. Google maps has been used to map businesses or crime locations, by being mixed with other data.
3. Use deep technology
Contextual spell-check is an example he gave. Google is able to use their vast data about spellings and all the wrong ways people spell things – to help users predict the right way to spell something.
4. Build for scale
"Dumb, unreliable, massively parallel, on lots of data."
Google does not try to make computers perfect. The most cost effective path for large computing situations is to use low-cost imperfect hardware, and make up for it with massive redundancy, running massively parallel.
5. Detect trends
Google Trends and Zeitgeist are examples. Trends can be detected using massive amounts of user data.
The Future
When asked about the future, he felt that today’s search technology will look primitive in 5 to 10 years. Verbal communications will continue to progress, where one day searching the Internet may just become talking to the Internet.
