Following up on last week’s Google I/O developer conference, Google has authored a new blog post highlighting “10 things you may have missed”—and they offer a strong indication of the company’s focus on A.I., mobile, and IoT technologies.
One major component of this is Google’s efforts in speech recognition, an area in which it has invested heavily as it’s viewed by Google as “[o]ne of the biggest uncracked nuts in A.I.” The company reports that more than 20 percent of its searches in the US are now done via voice; and given the major role that speech – and voice – recognition are expected to play in the emerging Internet of Things, it’s an investment that could have big payoffs further down the line.
Google says that a big part of its success in refining its speech recognition technology has been machine learning, and that too could play a major role in the IoT as a wealth of new data streams from connected devices is unlocked for big data analytics. Google’s Tensor Processing Unit (TPU), a newly-announced silicon chip designed specifically for compatibility with Google’s open source TensorFlow machine learning software, is another newly-announced development that could dramatically boost the development of A.I. solutions with applications in the IoT, mobile devices, wearables, and beyond.
Google’s researchers also see an opportunity to leverage a lot of this data available from connected devices—particularly smartphones—for a new style of user authentication that assesses metrics like geolocation, facial recognition, and even how a phone is held. That data can be used to produce a Trust Score, which the company aims to use to replace the password on its Android devices by the end of this year; and it’s easy to imagine that approach being applied to a number of other IoT devices in the future.
All of these efforts point to the growing wealth of data emerging from connected, smart devices, with Google seeking to fully exploit that data to the user’s advantage.
Source: Google Official Blog