Google's search engine based on machine learning is moving to support an integration with artificial intelligence (AI). Beginning in March, the entire search business will fall into the hands of John Giannandrea, an engineer who runs Google's research division. Here's why.
Historically it has been about typing a keyword into a search box and hitting return to query results, but for Google it has become equally about voice, biometrics, and a variety of connected devices such as the smart thermostat Nest -- not smartphones or tablets, but smart devices. The engine wants to serve you information based on your behavior and a variety of data inputs it receives throughout the day from a variety of connected devices. It must understand not only the advertising bids, but now must have the intelligence to connect with other devices to serve up much more information before you ask for it.
For years Google has been optimizing the price of the bid based on the quality of the content served to the person looking for answers. "It's not 'I pay $10 to get the top position,'" says Joelle Kaufman, head of marketing and partnerships at BloomReach. "Although I pay $10, the person bidding $5 might get the top space because their content is more appropriate for the searcher than mine."
That's machine learning, she says. Whether it's the ad unit or the great search results, Google maximizes relevance through machine learning. "Relevance is a mathematically calculated equation," she says. "It's what data scientists are trying to do to perfect search engines. Machine learning is the way you tune the relevance by using the results and the way people engage with the search queries to make adjustments by informing the algorithms."
Search will not remain in the engine. Time spent on mobile devices in 2015 exceeded the time spent on mobile devices. With this shift, the world will adapt to smarter mobile devices, not smartphones, and so will engines. It will transform smart mobile devices and wearables. Not mobile phones, but other types of devices.
In 2016, smart mobile devices will account for 44% of all mobile devices and connections, growing to 67% by 2020, according to Cisco. The number of wearable devices with embedded cellular connectivity will grow from 167 million in 2016 to 601 million in 2020.
Consider this. Some 39% of time spent on mobile is connected with Google product on either Android or iOS. Google owns 17 of the top 100 online properties of which six are in the top 10, according to Verto Analytics. Those properties include search, YouTube, Chrome, Google Maps, Google+ and Gmail. It also owns Google owns six of the top 10 most used smartphone properties--mobile browsing and apps--YouTube, Google Maps, Google Play Store, Chrome, Gmail, and Google Calendar.
So when you think about the future of search and why Google would appoint an AI and machine-learning expert to run search, remember the future.