Ok so for 3 years now I've been trying to get tickets to this event and finally got in! Even better, Phoebe got in this year too (must be lowering barriers to entry significantly in 2017 -- just kidding Pheebs!). Finally connected to WiFi. Here's a few photos of us and a pano of the main amphitheater.
Sundar is on stage! Giving a bunch of facts about history of I/O, advising everyone to wear sunscreen "plenty of sunscreen - it's on us - use it liberally" (google provided sunscreen in their welcome package -- real sweet, Google, now where's my new phone, google home, chromecast, etc.?).
7 products have over 1Billion users. Every day users watch over 1Billion videos on Youtube. Every day users navigate over 1Billion miles on Gmaps. 1.2Billion photos uploaded daily on Photos. Over 2Billion active devices of Android. Crazy scale.
Big shift occurring from Mobile first to AI first. Machine Learning utilized - for example, acceleration of use for smart reply in google assistant and gmail is 4x every couple months (shows that machine learning is getting better and better to the point that it's matching exactly what end-users want to say).
Google home now recognizes up to 6 distinct users based on voice. Google photos will automatically remove obstructions (example being a kid playing baseball and removing the chain fence in foreground).
Google Lens being unveiled today. Invoking google lens will allow you to point your phone and provides context and action - point phone at a wifi SSID and passcode and it will automatically connect you to that WiFi (pretty cool).
AI first datacenters. Tensor processing units (TPUs) - custom hardware for machine learning. They are coming to Google Compute Engine (Google Cloud) today. All Google AI will be under the umbrella of Google.ai - this is how Google plans to extend tools to developers.
Training neural nets to help with DNA sequencing, cancer identification, molecule discovery, or autodraw (as you're drawing google infers the image and you can select that image).
The most important product for AI/Machine Learning is search and assistant. Intro for Scott Huffman who leads Google Assistant.
Becoming more conversational, more available (offline and otherwise), and can do more. It should be the easiest way to accomplish tasks - conversation is a paradigm shift from touch/typing. Today we are adding the ability to type to your assistant on your phone as part of a conversation (switch between type, talk, and tap seamlessly in one conversation). Integration with Google Lens so assistant can see what you're looking at.
Google assistant now available on the iPhone (whooooooo caresssssss!?!?!)
Google assistant SDK allows any device manufacturer to incorporate google assistant. These products will have a badge that says "Google assistant built-in". It's rolling out on a bunch of new languages too.
It needs to be able to do more. Yes, of course it can search, but what can it actually *do*? Now it's open to third party developers who have built useful integrations. This is defined as "Actions on Google". You can ask apps certain apps to do certain things. Complete end-to-end solution for developers (payments, identity, receipts, etc.) platform handles everything. Here's a use case - As a new Panera customer no need to login, provide address, install an app, etc:
Google Home updates. Launched google home only 6 months ago. Added 15 new features since launch. Announcing support for proactive assistance. It will notify you of certain things - the example that he uses is having google home automatically talk to you to leave for an event *if* there's traffic - it knows what time the event is, what time it currently is, and that there's traffic so it will speak up and tell you that you need to leave early. Hands free calling is launching today (thank god, this has been a long time coming -- where is SMS though?!?!?). By default it calls using a private number but you can link to your account (it will dial out with your number when it detects your voice based on specific ). Free music service for Spotify (also adding support for soundcloud and deezer). HBO Now is now supported by Google Home. Visual Responses is being announced today. Google Home will provide visual response on the right device (e.g. TV, phone, etc.).
Google Photos updates. Launched 2 years ago. Automatically organize library and automatically curate photos for a given album (maybe a trip). 3 new features today.
1. Suggested sharing allows you to suggest sharing options based on who's in the photo, if the photos good, etc.
Adding a new sharing tab that tracks all shared photos and provides suggestions. This would be a great solution for trips.
2. Shared libraries - you can set up certain filters (only show certain people, only share photos after a certain date - example they used was after they met - that got some laughs from cheaters in the crowd, ha). Example: Anil takes two photos one with us and one with the girls and it will automatically save the photo of the girls to the wife's album as well.
3. Photo Books - this would be great for Karin. Easy to make, what used to take hours only takes minutes. All photos are already in photos since they have unlimited storage. You can use people search to search for photos of several people, then select photo book, then google photos automatically picks the best images for the book, lays them out, all you need to do is review make any tweaks and check out. In a few days the book shows up. Machine learning will automatically (based on your photos) propose new photo book suggestions based on your habits and photos. Pretty awesome.
All google I/O attendants receiving a free hard cover photo book (ok, but ummm can I get a new phone or home or something electronic???).
Integration of Google Lens with Google Photos. You can activate lens in photos and identify the buildings in the photo. Screenshot a friend sent you of a bike sharing business? Just activate lens, tap the number, and google lens will call the number.
Youtube updates. Incredible diversity of content in YouTube makes it special. 1Billion hours per day watched on youtube (crazy scale). You can interact directly with creators via comments, livestreaming, animated gifs, etc.
Living room updates. 360 video in the youtube app on the big screen. You can pan around the video while it's playing, kinda cool. This will be available live too - so you can interact with them as they are playing.
Superchat (what is this???) update. This is a feature that allows fans to interact with creators. Comes with a new API -- this allows fans to create interactions in real world. They show a demo of the product manager sending a "Superchat" to the SlowMo Guys (Youtube sensations, over 1Billion views on their channel) who are in the back of the amphitheater and that triggers a horn (automated) and then a bunch of balloons being thrown at them while they're live on Youtube. Meh, kinda gimmicky but I guess it's another way for 'fans' to interact with 'content providers' and it generates more revenue.
Android updates (ok I'm getting excited...). 2Billion active devices (that's just phones and tablets). Seeing lots of growth in Wear, Auto, and TV. Android O - developer preview launched 2 months ago. Expect a release later this summer. 2 major updates:
1. Fluid Experiences - here are some demos:
2. Vitals - keep vital system behavior in a healthy state. Battery life, stability, security, etc. 3 foundation building blocks
-Security Enhancements - use machine learning to continuously comb apps (scan over 50 billion apps on any given day). Google Play Protect (giving visibility) will come out of the box on every device
-OS Optimizations - on Pixel boot time is twice as fast. Apps run faster and smoother. Adding wise limits on usage and memory.
-Developer Tools - wouldn't it be cool if android could tell you what's causing battery drain, slow UI, and (something else.. didn't get it). Dashboard identifies these things and then provides recommended fixes. This is pretty huge.
Kotlin being added to Android O. Fully Android runtime compatible, interoperability with existing code. Beta release of O is available today.
How do we get devices to more people? Affordability in hardware, software has to be tuned for limited data connectivity and multilingual capability. New experience for entry level devices - Android Go - 3 features
1. Optimizes for low spec devices
2. Tuned android apps
3. Play updates
Seamless translation built right into the keyboard. Pretty awesome. As you type it'll automatically translate for you. All devices with 1GB or less will get the Android Go configuration and every release going forward will be supported.
VR and AR updates.
First VR. Daydream has a lot of items already supported. Standalone VR headset is coming. WorldSense allows VR to synch with movements. HTC is one of the producers, Lenovo is the other. Devices will be here later this year.
AR. Tango has been working since 2013, then released developer kit in 2015. The second device will ship this year. VPS (Visual Positioning Services) is coming soon. Pretty cool demo of going to get a specific screwdriver at Lowes:
AR is also very important for education and increasing interaction. Here's another demo:
Wrap-up with Sundar:
Google for Jobs. New initiative to connect employers and job seekers. Need to look at this to see if it's available in G Suite.
Anything on middle out compression?
ReplyDeleteHot swapping you-know-whats?
Delete