Today at Google I/O 2019, the company’s annual developer conference, CEO Sundar Pichai proclaimed that Google is working to “build a more helpful Google for everyone.” At face value this speak directly to newly announced features like Live Caption, a tool that will provide real-time captioning for audio or video content, allowing the deaf community to more easily engage with YouTube videos, podcasts, or messages from friends.
This mission is also realized in Google Lens, which gained myriad features including the ability to use the camera to read visual content aloud, both in the original language or translated into a dozen others. There is however, an underlying principle behind these technologies that benefits everyone equally. Google found a way to make these features possible while keeping the most sensitive personal data where it belongs, on the user’s device.
In today’s digital world where everyone is concerned with data privacy, and we are constantly reminded of the perils of massive data breaches like the ones that have plagued Facebook, it is more important than ever that companies find ways to ensure users have control over their personal information. Securing data is especially challenging when designing features that would normally require sending personal information to the cloud to be processed. This is why it is so noteworthy that Google found a way to shrink the Google Lens package down to just over 100kb, small enough to run on-device where it is easier to secure. Likewise, Live Caption was shrunk from around 2GB to 80MB.
As ProgrammableWeb highlighted in a piece titled “API Security is Easier Said Than Done”, very few organizations fully appreciate the difficulty in securing APIs. Which raises the question, if data is so hard to protect, is some data too sensitive to share at all. It seems that Google has realized that when it comes to the most sensitive data on user’s devices maybe the best solution is to just keep it on the device.