The key take away for me from Google I/O 2018 is their push of AI and that can mean big things in the war of Clouds with Amazon AWS and Microsoft Azure.
Google has long been at the forefront of innovation by perpetually improving and adding to their product suite, and now they are pushing AI into the consumer world with gusto.
AI and cloud computing are only going to get bigger, and as all the tech giants make a land grab for the cloud computing space, Google has taken a stance that they can add considerable value if you use Google Cloud by emphasising the accessibility of their AI capabilities.
"Kings of Data"
Google’s access to data through its product suite is unparalleled. Its investment in voice, accents, and associated video content via YouTube, and the popularity of Android that provides underlying voice and behaviour data has created real world context and experience to drive Google’s machine learning program.
AI, Machine Learning, and Natural Language Processing have already made waves in the corporate world through service automation, data insights and chatbots. But Google I/O showcased plans to bring AI to consumers in ways that have very visible and tangible benefits to everyday life.
Above all, Google’s revamped Assistant perfectly exemplifies this with its ability to make voice calls, and book appoints on your behalf is a huge leap forward, and ability for contextual conversations make this service natural, unjarring, and make this easy to see as a part of every home.
But as with all things Google, there is more than meets the eye, and these apparent value-adds for the consumer are also part of a bigger plan to expand its market share in Google Cloud.
"Soldier recruitment though ML Kit and TPUs"
Two other announcements from Google I/O tie this AI focus to their Google Cloud offering. ML Kit enables developers to leverage some of Google's data in pre-trained models for voice/image recognition, face detection and landmark labelling, which marks the start of making AI services for consumer much more accessible to everyone, and inviting for developers to take advantage of the power of Google’s data mine.
The other major announcement is generation 3 of the Tensor Processing Unit (TPU). At 8x faster than the preceding generation it competes directly with Nvidia’s GPU for AI modelling. And where do you get access to these advanced TPU’s? Google Cloud.
This is a perfect vision being realised by Google. They have the data required to build the most sophisticated AI models that exist today, they have a cloud platform that can provide access to these AI models, TPUs for processing power to build-your-own models, and they exemplify the value of using AI via their new consumer products like Google Assist, which in turn feeds more data into the engine. It’s perfect. As a developer, why wouldn’t you want to leverage this ecosystem for your new start-up or next generation product using AI, which will ensure Google’s infrastructure underpins these products for years to come.
Although Amazon, Microsoft and IBM all have AI offerings and cloud platforms, none of them can compete with Google’s access to and volume of data. Google are able connect the dots between pre-populated AI models, distribution platforms, and the processing power to bring AI to the consumer world. Google I/O 2018 was partly about demonstrating the potential in the consumer market, but the biggest result will be recruiting developers to build products on their platform, and turn the tables on Amazon for who is now playing catch up in the cloud computing space.