Google I/O 2015 live blog | CNET

Google I/O 2015 live blog

Google's annual developers conference is underway. Watch the live stream and follow CNET's live blog here.

  • James Martin 5/28/2015 5:21:27 PM
  • Tim Stevens 5/28/2015 5:21:28 PM
    "This is the beginning of a journey, just like we've done with Android for smartphones."
  • Tim Stevens 5/28/2015 5:21:38 PM
    Brillo goes into developer preview in Q3. Weave will be ready by Q4.
  • Richard Nieva 5/28/2015 5:21:44 PM
    Developers can expect Weave in Q4 2015
  • James Martin 5/28/2015 5:21:52 PM
  • Tim Stevens 5/28/2015 5:21:56 PM
    "We are very excited, and for the first time we are bringing a comprehensive end-to-end solution, and we hope we can connect devices in a seamless and intuitive way."
  • Tim Stevens 5/28/2015 5:22:17 PM
    "Now, we want to talk about how we as Google are improving the experience on a smartphone."
  • James Martin 5/28/2015 5:22:24 PM
  • Scott Stein 5/28/2015 5:22:31 PM
    How will all your devices know who you're speaking to? Will it be like a house full of fifteen kids?
  • Richard Nieva 5/28/2015 5:22:40 PM
    Pichai called back to Google's original mission of "organizing the world's information"
  • James Martin 5/28/2015 5:22:47 PM
  • Tim Stevens 5/28/2015 5:22:48 PM
    "If you're in a modern mobile phone, you can ask a question like 'What does Kermit's nephew look like,' and you get an answer instantly."
  • Tim Stevens 5/28/2015 5:23:00 PM
    We're now seeing that Kermit the Frog is Kermit la Rana in Spanish, too.
  • James Martin 5/28/2015 5:23:21 PM
  • Tim Stevens 5/28/2015 5:23:23 PM
    "What looked like a simple query, we understood voice, we're doing natural language processing... The reason we are able to do all of this is because of the investments we've made in machine learning."
  • Tim Stevens 5/28/2015 5:23:50 PM
    PIchai is saying how machine learning enables computers to know what a tree frog looks like, what he calls Deep Neural Nets.
  • Tim Stevens 5/28/2015 5:24:04 PM
    Neural Nets are nothing new, but they are incredibly difficult to train comprehensively.
  • James Martin 5/28/2015 5:24:08 PM
  • Tim Stevens 5/28/2015 5:24:18 PM
    "We have the best investment in machine learning over the past many years"
  • Tim Stevens 5/28/2015 5:24:29 PM
    Their word error rate has dropped from 23% to 8% since 2013.
  • Tim Stevens 5/28/2015 5:25:03 PM
    "This insight is what we will use to help organize user's photos, and you'll near about that in a few minutes..." ooh, a tease!
  • James Martin 5/28/2015 5:25:05 PM
  • Richard Nieva 5/28/2015 5:25:32 PM
    Now we're talking Google Now
  • Tim Stevens 5/28/2015 5:25:37 PM
    Now we're seeing how Google Inbox pulls all information about travel together. Google Now telling you when to leave, or showing your boarding pass at the airport.
  • Tim Stevens 5/28/2015 5:25:42 PM
    "In mobile, the need is even greater."
  • James Martin 5/28/2015 5:25:50 PM
  • Tim Stevens 5/28/2015 5:26:09 PM
    He's talking about the situation of someone messaging you about "The restaurant that I emailed you about" and how Google is going to try to figure that out, and save you having to find it.
  • Richard Nieva 5/28/2015 5:26:12 PM
    "We are working hard to be more assistive to users."
  • Scott Stein 5/28/2015 5:26:16 PM
    Google has some excellent speech recognition. And more efficient assistance is really needed for small stuff like watches.
  • Tim Stevens 5/28/2015 5:26:24 PM
    Aparna Chennapragada, Director of Google Now is now on stage.
  • Scott Stein 5/28/2015 5:26:46 PM
    This is making Google Now sound like my mom
  • Tim Stevens 5/28/2015 5:26:50 PM
    She's talking about how Google Now should remind you to call your roommate when you're visiting NY, book show tickets, etc. etc.
  • Richard Nieva 5/28/2015 5:26:54 PM
    "Your smartphone ought to be smarter." Chennapragada says.
  • Tim Stevens 5/28/2015 5:26:58 PM
    "We're working hard to find out how we can assist users in a mobile world."
  • James Martin 5/28/2015 5:27:16 PM
  • Tim Stevens 5/28/2015 5:27:18 PM
    She's pointing out examples, like the last train home in tokyo, an open house in your neighborhood, and a reminder about where your car is parked.
  • Tim Stevens 5/28/2015 5:27:22 PM
    That last one was for me, I think.
  • Tim Stevens 5/28/2015 5:27:32 PM
    To do this, the app needs to understand your context.
  • Tim Stevens 5/28/2015 5:27:38 PM
    Second, it has to bring answers proactively - without you asking for it.
  • Tim Stevens 5/28/2015 5:27:47 PM
    Finally, it has to help you take actions to improve the situation.
  • Richard Nieva 5/28/2015 5:27:49 PM
    Google now needs to understand your context, bring you answers proactively, and help you take action, she says.
  • Tim Stevens 5/28/2015 5:27:54 PM
    "In different contexts you need different things."
  • James Martin 5/28/2015 5:28:05 PM
  • Tim Stevens 5/28/2015 5:28:17 PM
    Disneyland, for example, vs "vegging out on a lazy Sunday" -- the app should act very differently.
  • Tim Stevens 5/28/2015 5:28:36 PM
    "We have built up a natural language processing engine, but we have also built up this powerful context engine, and we understand more than 100 million places."
Powered by ScribbleLive Content Marketing Software Platform
This week on CNET News