Mary Meeker’s Internet Trends Hint at Increasing Adoption of AI in Consumer Space
By Tony Vlismas
Approx. read: 5-10 mins.
There are surefire signs that spring is transitioning into summer: Google’s I/O, Apple’s Worldwide Developer’s Conference, and Mary Meeker’s Internet Trends Report. 2017 is no different. What has changed this year is the focus – amongst the consistent news of mobile growth and internet usage was data which points to the widespread adoption of artificial intelligence in the online world.
The series of slides focused on hands-free mobile usage hint at AI capabilities. Those based on advances in geo-tagging and improved contextual image searches show a certain level of supervised machine learning, but not necessarily AI. As the slides progress however, a deeper level of contextual learning and decision making is displayed.
Services like Snapchat can take stock of your surroundings when you use the service and serve you advertising based on items in your vicinity. Google can use your camera to provide advertisers information about how you came across their product in a store or otherwise.
However, the biggest revelation at the crossroads of internet trends and artificial intelligence is voice recognition. 70 percent of queries to Google Assistant were made using voice recognition software, well above the 2016 industry numbers (20 percent).
Amazon Echo numbers have risen to over 12 million units as of Q1 2017, and the number of skills and use cases it and Alexa have made available to its users is growing exponentially.
Still, the most staggering revelation in the whole document is Google’s success in machine learning and voice recognition. The word accuracy rate of Google’s efforts meets the threshold for human accuracy in 2017 (95 percent accuracy).
But what does this mean for the industry, and AI trends as a whole? One can hypothesize that as voice recognition becomes more intelligent, the frustration factor has dissolved. It is no longer a joke to speak to your Google Assistant or Siri and laugh at what they guess you dictated – their accuracy is breaking down the mental barrier of talking to your computer.
When accuracy is no longer a barrier to entry in computer dictation, and these devices acquire more useful skills alongside contextual awareness, voice recognition evolves from novelty to essential mobile feature. It’s a safe bet that next year we’ll see more than 6 slides from Ms Meeker around AI; with the release of more and more smart (or connected) home products in the months to come AI is a driving force amongst more internet trends.