Google's I/O developer conference kicked off this week and as in past years, it generated a lot of news spanning both consumer and enterprise-oriented scenarios (and of course, in some cases that line is a bit blurry). Here's a look at the top takeaways from the event's announcements for CXOs to consider.
Not Mobile First, AI First
Artificial intelligence has been the hottest trend in tech for some time now, and fittingly was the dominant focus of I/O. Put simply, Google wants to dominate the AI discussion and is making major moves to succeed in doing so.
It introduced Google.ai, which ties together all of its AI efforts in one place. It's aimed at both private companies, individual developers and academics and will focus on Google's AI research, tools and applied AI.
Google CEO Sundar Pichai unveiled a new project called AutoML, a neural network capable of designing neural networks. This notion has been a holy grail of sorts in the AI field, and Pichai says Google will make major strides on the relatively near horizon, writing in a blog post:
We hope AutoML will take an ability that a few PhDs have today and will make it possible in three to five years for hundreds of thousands of developers to design new neural nets for their particular needs.
AI is informing how Google evolves its products in a fundamental way, and the shift applies to the tech industry as a whole, Pichai added.
We are now witnessing a new shift in computing: the move from a mobile-first to an AI-first world. ... Think about Google Search: it was built on our ability to understand text in webpages. But now, thanks to advances in deep learning, we’re able to make images, photos and videos useful to people in a way they simply haven’t been before. Your camera can “see”; you can speak to your phone and get answers back—speech and vision are becoming as important to computing as the keyboard or multi-touch screens.
There is still a long way to go before we are truly an AI-first world, but the more we can work to democratize access to the technology—both in terms of the tools people can use and the way we apply it—the sooner everyone will benefit.
Google has its own commercial considerations for AI, of course, going beyond its core products. It is hoping to make Google Cloud Platform the go-to place for developing bespoke AI applications. Google's secret sauce for accomplishing that are its TPUs (Tensor Processing Units), specialized chips designed for machine learning workloads.
The first generation of TPUs were introduced last year, but focused on running Google's existing machine learning models more efficiently. Pichai annnounced that the second-generation of the chips, dubbed Cloud TPUs, will be offered through Google Compute Engine later this year. Cloud TPUs not only run existing models but can train new ones.
Google is also clustering the TPUs into what it calls "pods," which provide huge performance gains over past approaches. A new large-scale translation module once required a full day of training on 32 high-end GPUs, but accomplishes the same thing now "in an afternoon using just one eighth of a TPU pod," according to Google. TPUs will work in conjunction with TensorFlow, the machine learning framework Google open-sourced in 2015 to considerable success.
The company is running an alpha program for the TPUs and is also introducing the TensorFlow Research Cloud, which will make 1,000 TPUs available to researchers from both private industry and academia if they're willing to give back contributions to the open-source community.
Instant Apps Go GA
First announced at last year's I/O conference, Android Instant Apps are now out of preview and available to all developers. Instead of making users download and install an app, Instant Apps actually stream to devices from Google Play. Later on, users can decide to install them permanently.
Naturally, Instant Apps aren't as powerful as installed ones, which have deeper access to the device, but they do include useful capabilities such as payments and location.
Instant Apps provide a middle ground between websites and full-featured apps. That's a useful tool for enterprises to have in the toolbox, whether for internal users or outreach to customers. Constellation Research VP and principal analyst Holger Mueller noted earlier this year that Instant Apps have security advantages, since installed apps would involve MDM (mobile device management) issues.
The question now is how much momentum Google can build for Instant Apps out of the gate. Instant Apps capabilities will ship with Android O, the next version of the mobile OS, but will also be compatible with previous versions—a must, given the rampant fragmentation in the Android ecosystem.
Enterprises should take a look at how Instant Apps can fit into their overal mobility, marketing and internal IT strategies. Beyond the potential use cases, Instant Apps give IT leaders a new way to balance development resources; currently supported, full-blown apps could be replaced with lighter touch Instant Apps requiring less overhead for IT.
Google Steps Toward HR with Google for Jobs
There has been much speculation about which directions Google will head in the enterprise application market since the arrival of former VMWare head Diane Greene as SVP of cloud. While it's not clear that Greene's fingerprints are on it, a new Google service called Google for Jobs brings the company into the orbit of HR and HCM software. Pichai described the new service in a blog post:
[A]lmost half of U.S. employers say they still have issues filling open positions. Meanwhile, job seekers often don’t know there’s a job opening just around the corner from them, because the nature of job posts—high turnover, low traffic, inconsistency in job titles—have made them hard for search engines to classify. Through a new initiative, Google for Jobs, we hope to connect companies with potential employees, and help job seekers find new opportunities.
As part of this effort, we will be launching a new feature in Search in the coming weeks that helps people look for jobs across experience and wage levels—including jobs that have traditionally been much harder to search for and classify, like service and retail jobs.
Google has already worked with companies such as LinkedIn and Glassdoor to integrate them with Google for Jobs. What will be interesting to watch for are potential partnerships down the road with enterprise HR and HCM vendors.