July 9, 2012 | by Andrew Kameka
Android has supported voice search, voice commands, and text to speech features in some capacity for two years now. With the recently announced Google Now, Google has brought elements of all three sectors to form one sophisticated engine with an incredible deal of promise.
Google Now is not just another lame Siri for Android clone. Now is a complete voice input and text-to-speech system that taps into Google’s massive search database and Android’s interconnectivity. You don’t ask a silly question and wait 20 seconds for a reply, or repeat yourself incessantly because you run into server or system errors. You ask a question and get an answer, a proposition so simple that I can’t believe it’s taken this long for someone to get this close to being reliable. (Though that could change once Now becomes more widely available.)
At its core, Google Now is about search. That shouldn’t surprise anyone given Google’s backbone as a company, but much of the talk that I’ve seen so far about Now has regarded that it doesn’t have as many voice command features as Siri or even Samsung’s S Voice. All three apps are capable of searching for information, but Google Now doesn’t have as many second step features in the same way that Siri can find a restaurant on Yelp and book a table so seamlessly. Nor does it have some the device commands built into S-Voice on the Galaxy S III. It’s a valid praise of Apple’s search system, but it overlooks Google’s massive advantage in this field – Google Now isn’t limited by what it can do with its partners, it’s limited only by the knowledge users are willing to share.
Android Product Manager Hugo Barra revealed that Google Now isn’t just a fetch and deliver search engine – it’s a search tool that learns from your behavior and current position. Now can provide driving directions, but it also knows that based on the appointments in your calendar and current traffic conditions in the area, you’ll have to leave sooner to ensure on-time arrival. Flights that you’ve searched for, whether as a passenger or to pick-up a friend from the airport, may also appear with information on delays and terminals.
Now’s most intriguing feature is its ability to learn and adapt on a personal level. A user’s search history and profile information points Now to the right answers to increase the likeliness of relevance. For instance, a sports fan might ask a question about the Giants. Google knows that when I ask, I’m speaking of the NFL’s best franchise and reigning Super Bowl champs. When Bob asks, he’s talking about the baseball franchise for which he is a die-hard. Interestingly, when I ask “Did the Giants win?” in April, search knows that the NFL is not in season, so it presents a card detailing the baseball team’s most recent game. And on the off chance that’s not what I was looking for, a Google.com search is just a scroll away. I’d love to test out how it responds when both teams are in season, and the same goes for other similarly-named franchises like the Kings of the NBA and MLB.
Google Now isn’t just search, however. It can also use voice commands to send text messages or entire emails using voice typing. Yes, that’s the same voice typing that was upgraded to feature offline support and still has the most accurate speech to text engine I’ve seen from a mobile device. Other useful Google Now features include:
- Get exchange rates between currencies, including specific amounts (i.e. $100 USD = 79.466 euros)
- Get public transit directions and automatically know the next time a bus or train will arrive at the stop nearest you
- Call, text, or email your contacts with minimal button presses and voice dictation
- Set an alarms with voice commands
- Define a word and get links to Merriam-Webster or Dictionary.com for more information
- Be notified of heavy traffic on your daily commute, and then get an alternative route
The model for Google Now isn’t just about seeking information; it also delivers information without you needing to ask. Google can offer notifications when your favorite team is about to start playing, provide restaurant recommendations as you walk down a block, or even tell you what to order when you walk into that establishment to watch the game. It also reminds you about upcoming appointments and encourages you to get going if travel time may be affected.
This kind of interaction makes it clear that Google’s changes in privacy policies and its Knowledge Base search engine improvements have been all about pushing Google Now into being the unified search engine that knows its owner as well as its surroundings. Some users may not be too keen on that kind of data sharing, but there’s an option to disable location notifications.
Some device commands are not supported…yet
So much of the talk surrounding S-Voice and Siri is about them being “digital assistants,” but they are really just workers responding to commands. Google Now is aiming to be a step above that and be the kind of super assistant that knows your likes, needs, and wants without having to be reminded everyday. That small distinction is the difference between a gopher temp and the executive assistant who proves him/herself to be indispensable.
Google Now is still in its early stages and it’s by no means perfect. However, the gaps to fill are smaller than I’ve seen on similar products. S-Voice on the Samsung Galaxy S III is decent, but it’s neither fast nor reliable enough to become a consistent part of daily life. Combining Android’s voice technology with Google’s search power can make for a highly adaptive and intelligent relationship between the user and the device. Google Now is very good at the moment, and I can’t wait to see it be great.