Five future AI applications
Dr John Zakos’ mission is to humanise artificial intelligence. The Program Director of Virtual Agents at IBM and Honorary Professor at the UTS Faculty of Engineering and IT shares his thoughts on how AI could be used in the future.
Just imagine: groceries just turning up at your place. You don’t have to go online to order anything. Imagine them turning up. Like avocados when you get home. Tomorrow you’ve got toilet paper, or bottles of water.
This is an AI system that, in the background, knows who you are. Maybe it has input from your doctor about your health profile and what you should be eating. Or input from your parents, family and friends on how often you eat out, your social habits. That’s artificial intelligence learning your preferences. You’ll never have to walk into a supermarket ever again.
Blood biopsies for advanced illness detection
Liquid biopsies could do super-early detection and diagnosis of illnesses, especially cancer. Today, you usually have some symptoms, maybe pain in your body. And you go to the doctor, get scans. The doctors come to a conclusion, and you’re diagnosed with something.
But imagine that you go for a blood test. Before you even have any symptoms, a system can tell if you have cancer or not by analysing your DNA profile, the floating DNA in your blood. Here we have medical technology and processes, combined with core AI algorithms, that look over the three billion dimensions of DNA and tell us whether we have cancer or not. And by early diagnosis, allows prevention or therapies to be introduced faster.
If you look at some research that’s come out earlier this year, it’s starting to happen. In a few years’ time, it might be reality in terms of a real-life product.
Personalised self-driving vehicle rental—for free
This is a little obvious. Self-driving vehicles—cars, planes, you hear Uber and others talking about it. But imagine you get in a self-driving vehicle that is free because it is paid for by advertising. Not only that; it’s a vehicle that knows you: what temperature you like the air conditioning; if you like to fly at a high or low altitude. It knows what music you like played in the background, and so forth.
So it’s beyond self-driving cars. It’s really about personalised intelligence that knows who you are— the underlying theme of all these applications.
Virtual personal assistants
This is more straight forward: a [virtual] PA that performs tasks for you. Google demonstrated Google Assistant making a phone call for a hair salon reservation. But imagine a PA doing all sorts of tasks for you. Interacting through multiple channels: phone calls, emails, SMS… paying your bills and so forth.
And imagine if it could just do a small portion of those tasks really well and how much of a load it would take off you. Once again, it’s about knowing who you are; understanding you so well that it goes and pays a bill tomorrow, or makes a booking at a hair salon or a restaurant you really enjoy.
Some of the work that we did at Cognea [the AI company Zakos co-founded and acquired by IBM] was in intelligent personal assistants and there’s a lot of ongoing work happening in this area, so it’s definitely part of the future.
In Japan, people are subscribing to a robot like Pepper that comes into their home. This is one dimension of what AI offers: emotional robots.
AI is not just about cutting costs and doing things for you. A big part of life is feeling good, feeling well, growing and having fun. So I just want to highlight the emotional aspect of artificial intelligence. Whether it’s an enterprise business or consumer application, emotions have a big role to play—the more emotional we can be interacting with technology, the better the experience.
Baby X is a [virtual] baby who makes noises and expressions. This is made by some colleagues at a University in New Zealand. The point I’m trying to convey is having emotions in an interaction is compelling in a good way. The technology exists and I think it plays a very relevant role.