Paul Yarin

Paul Yarin

By Tara Chklovski

In light of Iridescent’s breakthrough initiative to make Artificial Intelligence learning accessible to underserved communities, I sat down with Paul Yarin to get his thoughts on how communities can innovate with AI. Paul is Director and Head of Technology at global aerospace company Thales xPlor, in addition to being an entrepreneur, designer, and inventor. He has spent thousands of hours thinking about good design and creating innovative, beautiful technology products.

The day before I interviewed Paul, Reuters broke news that shares for NVIDIA hit a record high, thanks in part to the rapid implementation of the Volta chip. NVIDIA claims that this chip could eradicate cancer and power AI systems, such as driverless vehicles that may reduce car deaths. Positive news like this about AI can be few and far between. Through our conversation, Paul and I highlighted the need for a positive dialogue surrounding AI, as our children’s generation will use it to solve tomorrow’s toughest problems.

Tara: What inspires you?

Paul Yarin: I am inspired by complicated problems and people who successfully navigate the complexity of the world. It could be a technical product, company, or movement in the arts. How did electronic dance music become such a thing? Or the electric guitar and rock ‘n’ roll? As a collection of humans, we are adaptable, curious, and have a hunger for trying something in a different way and a faith that there is a better way.

What makes a good problem?

A good problem is one that a lot of people want to solve and with new solutions that can be brought to bear over and over again. For example, we have a challenge in caring for older people; this is not a new problem. But it’s still something that technology can be used for. And because technology is changing very quickly, every generation brings a new set of tools to this same problem.

Caring for younger people and helping them develop is obviously a good problem, too. How do you help young people? How do you give them enough to be challenged and not so much that you are crushing them into a conformist pattern? How do you give them enough freedom to reward their curiosity but enough control so that they don’t hurt each other, themselves, or you? How do we stay healthy is always a long-term problem. How do we stay happy is an even more fundamental problem!

What does achieving and keeping happiness mean to you?

There is nothing worth doing that doesn’t bring you frustration and irritation, at times. But happiness is when you believe in the significance of what you are doing and you manage to keep going!

An important component of Iridescent’s Curiosity Machine AI Family Challenge is that families will try to solve a problem in their communities, with AI. What advice would you give them about finding a good problem?

Finding problems is not as easy as it might seem. You have to reframe your thinking when you look at the world. One way is to accompany somebody who is different from yourself—somebody older or younger, from a different culture—and make yourself a fly on the wall, observing the world the way that they see it. Don’t try to change their world, don’t try to interact with it too much; just try to see the world through their eyes. For example, if you hang around an older person you may find that while it may be very easy for you to [climb] the steps into a building, it can be a complicated problem to solve for the older person, who has limited mobility.

Great answer. We want families to solve these kinds of problems with AI. What aspects of AI excites you?

A lot of them! Computational psychiatry is interesting to me. The idea of analyzing people’s behavior in a more objective and systematic way by looking at the words they use and the timing and inflection of their speech is a pretty interesting domain of AI.

Essentially, what is attractive about computing is that it’s never tired, it’s always available, it’s always there, and it’s incredibly cheap. The network connectivity and gadgets that you carry around keep getting more powerful [and stay] at the same price. This doesn’t get talked about so much, but it is very remarkable. There’s no reason why AI shouldn’t be affordable to a really broad population—even to the poorest group.

How do you think AI will help strengthen society and communities?

Well, the strength of a society comes from the values of its inhabitants and their resources such as economic and human resources. Humans are fallible in a million ways and so we’ve created human organizations to deal with the variation between people. For example, we have a jury in a court because any one person might have a bad day, or another may have extreme opinions. But, you are more likely to get a fair outcome if you average opinions across all 12 people. The interesting thing about computers is that it is just as easy to get 12, 12,000, or 12 million AIs on something. It is not restricted by the same availability of scale that humans are. I can imagine a day when AI may come up with defensible decisions in a court of law that are actually more consistent and more fair than what humans come up with. That would strengthen society.

What do you think is the best way for children and families to learn more about AI?

I think the first thing is to separate yourself from some of the misinformation about AI. There’s a lot of science fiction and popular fiction about it, and some of that may come true! But you need to step away from the drama to understand what is actually feasible and possible today.

AI is in a lot of products today and will continue to be in consumer products. We have conversation agents like OK Google and Alexa that can do natural language processing. We see more and more image processing via Google Image searches, and we will soon have autonomous cars that will do even more sophisticated image processing to decide how to drive. We have AI algorithms that make decisions as to which movies and videos [are best] to show you. In some ways, the most accessible way to learn about AI is to play with these products that are around you every day.

Beginning in 2018, children ages 8 to 15 and their families will be able to join the first global initiative of its kind — a program for families to learn core concepts about Artificial Intelligence and use these tools to solve problems in their communities. Learn how you can take part.

This post is the first in our AI in your Community Series