I remember once when I was much younger, doing some online clothes shopping. I know this seems like a fairly routine thing to do, however the reason I remember this moment so vividly was the disbelief, ten minutes later whilst reading through an article on a football website when I saw the EXACT items I’d just been looking at, linked in an advertisement down the side of the page. At that age, I had no idea of cookies, or any other tracking devices that allowed advertising to be specifically tailored to the type of content you viewed. Today, before writing this piece, I had the exact same thing happen to me, and smiled at the memory of once being so amazed at what is now something that goes almost unnoticed to me. Ironically, the ad appeared on an article I was reading about Google’s new AI tool that turns your selfies into custom emojis. Google is leading the way in AI & Machine Learning, and is continuously experimenting with different ways they can integrate AI and ML into our everyday experiences. For example, Allo — Google’s most advanced messaging app — has an integrated AI assistant which can perform many useful tasks. For example, it can suggest a personalised quick response to a message you have just received - created using data gathered about writing style, and phrases you commonly use. To me, that’s awesome. We spend less time thinking and typing responses, making us more efficient in our communication. Some of the features available to Allo users However, many people are still wary of AI; concerned about how much this ‘machine’ knows and collects. But to me, this viewpoint is flawed, as it would appear as though these ‘machines’ already have access to more information that most people realise. Consider briefly, the definition of technology; “the application of scientific knowledge for practical purposes, especially in industry”. Technology has always been — from the invention of the steam engine, to development of self driving cars — about improving the efficiency and/or effectiveness of performing a certain task. With this purpose in mind, it excites me the way that Google is using the data it gathers to create tools that allow users to be more efficient in their interactions. Sure, Allo’s newfound ability to turn your selfies into emojis hardly leads to greater productivity. In fact I can see myself spending too much time making emoji’s — resulting in a loss of productivity! But for a user, a machine scanning your facial features and converting it into a custom designed set of emojis, makes for a much more personal, and far more interesting UX. Google’s selfie-come-emoji tool is a sign of the company’s desire to integrate AI and ML to improve the UX in their applications. However, I also know that privacy advocates slam Allo for not using end-to-end encryption as a default. Our conversations are read not only by people in the chat, but also by Google’s Assistant. This means that all our conversation data, is stored in Google’s servers, and thus, only protected by the security of Google’s servers. Obviously, the very act of storing this data, means that it is vulnerable, to a certain degree, to an attack. So yes, as much as I like the features that Allo has, of course I am concerned about my privacy. To summarise, my point is this; Allo uses AI and ML for the primary purpose of technology; to create a more efficient way of completing a task, which is in this case; communication. Not only that but it also includes other features, such as the selfie-come-emoji tool, that contributes to a better UX. However, in order to utilise this functionality all our conversations must be read and collected by Google’s Assistant, meaning we are potentially jeopardising our privacy in doing so. So I leave you with a few questions, which I’d love to hear your thoughts on. Tweet me: or share your thoughts a response below. @nicholasrlane Are you willing to potentially jeopardise the privacy of your conversations, in exchange for features that improve the productivity and efficiency of your necessary interaction? Are features that create a greater UX, worth it for the potential loss of privacy? And finally, truly how vulnerable is the data that Allo collects to an attack?
