Pixel launch: Is Google Assistant's AI an assault on privacy?
Sure, artificial intelligence can make life simpler by suggesting you places to eat, tell you weather, or even crack a joke. The more you use it, the more it learns about your habits and preferences. The protocols pertaining to artificial intelligence, or the lack of it, make our data less secure and more vulnerable.
Late Tuesday, Google showcased how the combination of Pixel smartphones, Google Assistant and other accessories can make life simpler. The Internet giant is betting on people’s hunger for technology with its artificial intelligence-powered Google Assistant.
Unlike its predecessor Google Now which can set alarms, check your mails, help you with navigation and web searches, among many other things, Google Assistant is highly personal. The artificial intelligence-powered assistant combines machine learning and image recognition.
Imagine this: You are chatting with a friend about a movie you both want to watch. Behind the scenes, Google Assistant is tabulating a vast number of metrics such your location, preferences, cineplexes you like and personal thoughts to throw up the best possible option. This tabulation requires the AI to learn about you (read collect and analyse data) as much as possible to give you an accurate results.
Sure, it can make life simpler by suggesting you places to eat, tell you weather, or even crack a joke. But what is it asking in return for the highly-personalised service? Answer: Data access and personal information. The more you use it, the more it learns about your habits and preferences. The protocols pertaining to artificial intelligence, or the lack of it, make our data less secure and more vulnerable.
Unencrypted messages are required for Google Assistant or any AI to function to the best of its capabilities. This unfettered access to data puts users at risk both to government surveillance and hackers. Not just that, Google has responsibilities towards its partners too and this can prompt the company to offered sponsored results by possibly creating a revenue stream. Social media sites Facebook and Twitter are already offering sponsored searches in users’ feeds.
Meanwhile, Google’s chat app Allo has a very good encryption. Turning on encryption would mean that you won’t be able to use Google Assistant, which defeats the purpose of artificial intelligence in the first place.
For now users are stuck between a rock and a hard place. They can utilise the potential of AI for convenience — exactly Google’s agenda. Users should understand what they stand to gain or lose by sacrificing encryption for AI-led assistance.