Rand Hindi wants access to your phone. Every part of it. He wants your emails, your browsing history, your GPS tracking. Why? Because this revolutionary darling of the French tech sector wants to create a personalised AI that will learn from you and assist you in your every day life.
If you have any concerns regarding the access of your data and the theft thereof, he acknowledges those worries as such: “There is this narrative that you have to choose between AI and privacy, which I think is, honestly, complete bullshit.”
Hindi’s aim? To reduce the need for humans to interact with their own technology. “If you can have an (AI) assistant that’s smart enough, that understands everything about your life, then it can start simplifying the interaction you have with those devices and even someday automate most of them.”
Hindi argues that machine learning would allow for humans to lead more fulfilling lives, to stop reaching for their phones during dinner (or sex, as he ruefully admits in a TED talk). Through machine learning algorithms, humans will look at their own phones less, and head back to a more fulfilling state of living in the natural world.
That’s right, folks. We got a live one.
Caged by machines of loving grace
Whether you are a technophile or phobe, Hindi’s core argument concerning the negative impact of interconnectivity rings true. “Back in the 1990s, we didn’t have any connected devices; this is what we call the unplugged era. By 2005, it was three devices on average per person. The issue is that those devices actually do not talk to each other and do not really understand what is going on. Which means that you, as a user, need to interact with exponentially more technology than what you used to, wasting even more time than what you should in practice,” says Hindi.
Hindi’s aim? To reduce the need for humans to interact with their own technology.”
Anyone who has owned a mobile phone has felt or heard their phone vibrate, only to find the phone has not received a message. This well-documented ‘phantom sensation’ is Pavlovian, a learned response to the pleasurable sensation of receiving a communication from someone you love. It is a symptom of our deep interconnectedness with our technology.
Hindi argues that using machine learning can allow us to utilise our tech more efficiently as we progress into an era of ‘ubiquitous computing’. “I do believe that we need to use AI to make technology disappear into the background, so that we can start doing things that we actually care about and not waste time just trying to make your phone do some kind of very simple action,” he says.
But how does that work?
Hindi’s start-up Snips was founded in 2013 as a research lab focused on building new interfaces between people and machines using machine learning. In his presentation for With The Best, he provides an ingenious example of this interfacing, citing the need for a location service to follow your movements when you get on an underground subway.
Using a mobile phone’s inbuilt barometer (by the way, we’ve had barometers in mobile phones since 2011), Hindi can co-ordinate your movements with your phone’s GPS locator to track your movements underground.
At Snips, we actually don’t have any of your data, we don’t want your data, we consider that to be a toxic asset.”
“Every time the train accelerates in a tunnel, that creates a physical effect that generates a pressure variance. When you measure the peaks between two consecutive pressure events, and look at the time that separates the two, you are basically able to determine which station you have being going through,” he says.
Hindi cites other examples: using Android’s ability to capture and analyze the content of a phone’s screen allows for an algorithm to scan your applications for contextual clues that will assist the other apps to be more efficient.
If you want to find an Italian joint near your Airbnb reservation, there’s no need to turn on location services and manually search for it. Your Airbnb reservation has your new address stored, therefore your AI assistant can extrapolate where you are, begin a search for restaurants in the area and match it to other Italian restaurants that you’ve enjoyed before.
And if you want to call an Uber home, the reservation will be made with the data from your reservation rather than the sometimes-inaccurate GPS locator. What Hindi wants to simulate is a user to interface with the user interface.
Dude? Where’s my privacy?
“The more access to your life your assistant has, the more intelligent it can become. If someone hacks into my assistant they basically pretty much have access to everything, they can lock me out of my own life, so that would suck,” says Hindi. This is quite an understatement. But Hindi considers the question of piracy and data theft a central tenet of the AI assistant. Trust is important for establishing a consumer base, but Hindi’s goal has been to move the machine learning algorithm away from the cloud and localise the service on your handset.
“From the beginning, the strategy we’ve been adopting has been to run every single thing on device on both iOS and Android. There is nothing being sent to the cloud. At Snips, we actually don’t have any of your data, we don’t want your data, we consider that to be a toxic asset,” says Hindi.
It is a practical sentiment from an entrepreneur whose vision and goals are utopian. In many ways, Hindi is beginning the real work of the 21st Century, an era unprecedented in human history. It’s been a mere 25 years since human beings lived in the ‘unplugged age’, but Hindi is confident that it’s only a matter of time until the tech for AI assistants will be ready. He thinks it will change the world.
“Since we were kids, people have been telling us that the future will look like a scene from The Fifth Element; police cars everywhere, huge-ass up-in-your-face metal buildings, very oppressive, very stressful. It doesn’t have to be like that, because when you talk to people most of them want to be in the world, they want to actually feel at peace, they want to feel this connected, they want to have time to do their own things. To me, AI and privacy and design are the three key components to make technology disappear, so that we could actually have time to do things we enjoy again.”
Originally posted on BeMyApp Media This article was based on Rand Hindi’s talk for With The Best’s web conference on AI.