Google I/O begins on May 17th, 2017 at 10AM PT / 1PM ET with CEO Sundar Pichai taking the stage for the keynote. Held at the Shoreline Amphitheater in Mountain View, CA, there will be one main keynote, a developer keynote that same day, then events running throughout the rest of the week.
Under Pichai, Google has developed a new cadence for its product announcements. It's not quite as rigidly predictable as Apple’s, but it's getting there. So for the annual developer conference, Google I/O, we can be fairly confident in a few things: there won't be a new phone, machine learning will get plenty of stage time, there will be more details on the next version of Android, and there will be some weird surprises.
If you're wondering what the rest of that yearly cadence looks like and how Google I/O fits into it, here's a quick primer:
- March: Google Cloud Next, a conference focusing on Google's cloud services and enterprise services
- March: For two years running, Google has released a very early developer preview of Android
- May: Google I/O
- October(ish): New hardware (last year we got the Pixel)
That schedule doesn't tell us exactly what to expect from I/O, though it does mean that we can bet on a handful of beats that matter to Google's developer community. But don't think that a consistent timeline is the same thing as focus: the I/O keynote often feels as scattershot as Google itself. There will likely be lots of different divisions getting stage time to talk about their products, and keeping up with them all will be a marathon-length sprint. Fortunately for you, we'll have a live blog with myself, Casey Newton, Adi Robertson, and Vlad Savov to explain everything in real time.
When we try to predict what to expect from I/O this year, we end up looking at a mix of sure things and wild speculation. So let's get into all of it!
Google Assistant, Home, and AI
It's a sure thing that the Google Assistant will get a lot of attention this year. Firstly, Google is still making a push to improve its capabilities in the Google Home speaker while also making sure it works on a ton of other devices. It is also tied into one of Google's core strengths: powerful machine learning algorithms.
Basically, Google Assistant is the product that touches nearly everything Google makes and is the most obvious, direct example of Google's AI chops. It's also generally regarded as better than its competitors Siri, Alexa, and Cortana. Google will want to show it off.
But that lead isn't as strong today as it was a couple of years ago: Alexa's wild and wooly developer ecosystem means that the Echo talks to more software and devices than Google Assistant. Apple can (and often does) make a strong case that Siri does a better job of protecting your privacy than Google. And Microsoft isn't sitting idly by either, as Cortana-powered home speakers are coming soon. Heck, even Samsung is getting ready to take a run at Google with Bixby on the Galaxy S8.
So for Google, the pressure on the Google Assistant to maintain (or widen) its lead is very real. What exactly that looks like for Google isn't totally clear, though. A lot of the innovations we've seen on previous iterations of Google's consumer AI haven't really achieved their potential: Google Now is mostly ignored by the majority of people and Now On Tap has been relegated to a single button almost nobody presses inside the Assistant.
Chances are we'll see a mix of new features, new devices that can run the Assistant, and updated developer tools. Pay special attention to that last one: Google finds itself behind Amazon in terms of mindshare for home speakers that "do things," so it needs to close the gap there.
We should also get an update on Google Home and its capabilities — again, Amazon is pushing hard on the home speaker front. And don’t forget about TensorFlow, Google’s machine learning platform, it’s likely to get a lot of stage time.
As we had last year with Nougat, we've already got an early look at the next version of Android, O, in the form of a developer preview. And as we saw last year, the developer preview only has some of what to expect. At Google I/O, we're sure to find out more details of what's coming in Android O.
Will we find out the full name? Last year Google held a contest to announce the name. This year it may do the same even though the conventional wisdom choice of "Oreo" feels like a fait accompli. (Or, given the dessert theme, a parfait accompli. Sorry not sorry.)
But first, here's what we know about O so far:
- Snoozing notifications: You'll be able to half-swipe notifications and hit a "snooze" button, so they'll appear later, when you actually want to deal with them.
- More granular notification control: Apps will be better able to "group" their notifications so you will be able to block just the ones you don't want without having to hunt through app settings.
- Improved battery life: It seems like Google promises this every year. And this year, it's doing it by more strongly limiting what apps are allowed to do in the background. That's a little more like the way iOS handles its multitasking.
- Better Bluetooth: Sony has chipped in some code to improve Android's Bluetooth audio performance.
- Cleaner settings: Whether you're Google or Samsung or LG or whatever, Settings is your favorite app to mess with and change every year. Because trying to make sense of Settings is awful. In O, the main page will be simpler and cleaner, hopefully helping you find what you actually want faster.
- Change buttons: Deep in the System UI Tuner, there are options to mess with the core trio of Android buttons at the bottom of every phone, as well as options to change the quick launch apps on the lock screen.
- Improved support for physical keyboards: Official support for Android apps is coming to Chrome OS eventually, and when it does you'll want to be able to use Android apps with your Chromebook's keyboard more easily. Done and dusted.
- Picture-in-Picture video: Finally.
- Theming options: Maybe! We know that Android O will allow for differently shaped icons, a core part of most theming. There are some other signs theming could happen.
Take all that together and you're looking at a pretty healthy maintenance release, a "let's fix a bunch of stuff under the hood" update. The question is whether Google will be satisfied with that or if it will want to do something a little more ambitious. Last year, we saw similar feature bullets in the preview and then had Daydream VR unveiled at Google I/O. This year? There haven't been very strong rumors about something that big, but maybe Google will surprise us.
Android apps on Chrome OS
Android apps work on Chrome OS right now, but they're in beta and they don't run super well. I've heard that they were originally developed on the Android M platform and the end goal is to release Android on Chrome OS on the same cycle as Android on phones and tablets. So in a worst-case scenario we might be waiting until O is official. Even if Google has decided that it needs to take a little more time to get Android on Chrome OS right before making it official, it has to give us some sort of update on the project at Google I/O. The anticipation and demand for this sort of cross-compatibility between devices and platforms is only going to expand. So expect that update, and if we're lucky expect an announcement of an official release date.
When that time comes, we're assuming that the windows housing Android apps will be freely resizable and fully capable — at least as good as they are on Samsung DeX. I'd also like to see Google clean up the app launcher so it's easier to distinguish between Android apps and Chrome apps. Beyond that, the big challenge will be getting third-party developers making apps that work well on large screens.
VR and AR
It's been a full year since Google unveiled its VR platform, Daydream. In that year we've seen only a handful of non-Pixel phones updated to support it. That's not great. Google is facing serious competition in the VR space — and it hasn't really had anything notable to say about augmented reality in a very long time.
So at minimum we should expect an update on what's going on and what's coming next for Daydream. It might be too much to hope for a standalone headset or true inside-out VR tracking with Project Tango, but nobody would really complain if Google managed to get it done.
Instant Apps, Chrome, and the web
It's okay if you don't remember "Instant Apps," because beyond announcing them at I/O last year it doesn't seem like Google has done much with them. The idea is that instead of installing an app before you run it, an Instant App can start right away and stream itself to you over the internet. Then it goes away when you're done, just like a web site.
Instant Apps are a cool technology, but whether they're truly viable for developers is entirely another story. Because they require specific pieces of Android to run, they're not likely to be as widely available (nor widely adopted by developers) as more traditional apps or good old HTML apps.
But they're on the agenda this year at I/O, so presumably Google is getting closer to making them officially available on a wider basis. It's neat tech, and worth keeping an eye on this year.
As for the web, Google still cares about it a lot even though things like Instant Apps, Chrome-only web apps, and even AMP tend to undercut it. Google will surely show off more features available for Progressive Web Apps and AMP pages.
For Chrome, keep an eye out for anything the presenters might say about battery life and Windows 10 S availability. Microsoft has been throwing down a lot of gauntlets lately, and it would be weird if Google didn't respond.
Allo, Duo, and Hangouts updates
Allo has seen relentless upgrading since it was announced last year, but a messaging app from Google is rightly judged by how many users it has — and Allo doesn't have very many. For I/O this year, I would not anticipate that Google will announce SMS integration or anything close to how iMessage works. The company is just too nervous about pissing off its carrier partners who are trying to get everybody on next-generation RCS messaging.
Instead, we should get an overview of all the features that have been added to Allo and hopefully a preview (if not a release) of the desktop app and multi-device support. That desktop app is likely to work the same way WhatsApp's and Telegram’s do: on the web and tied to your phone for its key operation.
Duo, meanwhile, seems to be a small hit. We may see some updates, but this is more of an "if it ain't broke" story. Hangouts just got a massive overhaul to become an enterprise-focused Slack competitor. We may see it demoed and may get a release date, but don't hold out hope that it's ever going to be good for consumers again.
We love to joke about Google's messaging strategy, but outside of some jaunty experiments, it's actually pretty simple:
- Allo and Duo for consumer messaging
- RCS support for carriers
- Hangouts for businesses
The rest: Android Wear, Android TV, Chromecast, and more
Many of these products have had recent updates and though they all have things that should be changed or fixed, I suspect that we're not on the right cycle to get them this May. Android TV, in particular, is basically Google's White Whale and if the company keeps going out to sea with random attempts it's going to be embarrassing. Instead, we should see some small, developer-focused announcements and perhaps an update for Chromecast.
Beyond all that, don't forget that this is a developer conference, which means we'll get updates for all sorts of developer tools like Firebase, Google's Cloud tools, and Android Studio.
What else? Who knows! Google’s gonna Google. It should be a fun keynote.