Trailblazers in AI: An Interview with Michael Kanaan
Michael Kanaan is a member of Forbes 30 under 30 and was formerly the co-chair of Artificial Intelligence for the US Air Force. He is the Director of Operations with a new partnership at MIT’s premiere AI lab: CSAIL; and the author of a new book: “T-Minus AI”. The book offers an intriguing look into the different disciplines of AI, and how they are already shaping our present/future.
Key Points:
- Get the full picture: AI is often falsely depicted in the media as humanoid robots with the sole mission to end the world and take our jobs. Well, as scary (or cool?) as that sounds, let’s make sure we realize that it’s much more than that!
- AI is not limited to STEM: Throw away the narrative that “only STEM backgrounds need apply” to work in AI. Michael explained how he in fact is a Business major, and that coming from different backgrounds allows us to gain unique perspectives in a team.
- It’s easy to get involved: In this new digital age, it’s easier now more than ever to get involved in the AI space. There are countless communities to join and loads of online resources to up your knowledge on AI.
How would you describe what Artificial Intelligence is?
Michael: When we talk about AI, some people tend to imagine the Terminator. For others, it might be Baymax from “Big Hero 6”, “WALL-E”, or Hal from “2001: A Space Odyssey”.
What those examples all have in common is the assumption that AI is unavoidably destined (sooner or later) to develop it’s own consciousness and autonomous, evil intent. But when it comes to these portrayals of AI, too often they generate an array of fears by focusing our attention on distant and somewhat dystopian possibilities, rather than the present day realities. They usually depict AI as an alignment of computer intelligence with consciousness, but then frighten us by portraying a world where it’s not only conscious, but also evil minded and self-motivated to overtake and destroy us.
However, I think it’s better to talk about AI in a little bit of a different way. Think about this: Do you ever wonder why you spend so much time on the app “TikTok”? You’re constantly finding new content, dances, people, videos, or just music you love. Well, that’s because every time you’re on there, the app is tracking those hours and hours of activity and creating recommendations. And…that’s AI! You can totally feel the power of it just through this example. It’s why you constantly get the content you love across so many different platforms.
As most people know, AI is essentially nothing more than an algorithm or piece of code. It’s designed to evaluate data for the purpose of drawing information in a way humans may not be able to perceive. However, it’s important for us to know the field of AI, the ways we go about creating it, and how it should be used. AI really is the electricity of the 21st century.
For many, there may be a belief that in order to really understand or be involved with AI, you have to be a programmer in Silicon Valley or an Ivy League graduate who’s an ace at math. Is engagement in this technology really limited only to those within STEM fields?
Michael: Well, I myself am actually not a “STEM” person. We need to change this connotation. As a Business major, I wasn’t traditionally educated or trained in that way in the typical university setting. It’s not about just being a coder as the only way to contribute too. Frankly, I think the future rockstars in this space — that we’re currently seeing in Silicon Valley — are the philosophers, ethicists, writers, and future politicians that look at the outcomes of artificial intelligence.
Inevitably, we’re going to have people writing code. But figuring out what to do with it and coming up with ideas for business, as well as broader society, will be what’s most important. There’s actually an analogy I really love using from Cassie Kozyrkov — the Head of Decision Science at Google. She asks the question: “Do you know how to build a microwave?” Most people would say no, but there’s not many who haven’t used a microwave. The real question we should ask is: Why do you trust that microwave? If you can’t explain it, why would you trust it?
Just like how there are people who aren’t classically trained as electrical engineers who do phenomenal things with electricity and build products, there are people who make food that’s meant to go in microwaves. We all know not to put tin foil inside of it. I think that blending the hardcore engineering work by brilliant minds with how we think about AI is what’s really important. That’s why you don’t have to be a STEM person to work in the field.
Although many current students may not work directly as AI scientists, they’re curious as to how they could be using AI in the future. What are some examples of AI tools or applications they could potentially be using, even if they may not necessarily be the developers behind the underlying algorithm or model?
Michael: I could name off a few fun examples. One that I’ve seen off the top of my head is an app where you take a picture with your phone, and it’ll identify a food like Sushi and the ingredients in it. The purposes of that would be to detect what your food contains if you have allergies, or if you simply want to know what kind of fish is on your plate.
Another similar example is used to check math problems. Basically, you take a picture with your phone of a math equation, and then it tells you whether your answer is correct or not. We’re also seeing it move into everyday life with video games through using AI for ray tracing — which is a way that light bounces off material. So, we’re using AI to predict how that behavior is and that’s why graphics look so good right now. Additionally, we’re using AI in the arts. There’s AI that makes paintings, creates music, or writes poetry.
Essentially, we’re taking this technology that we’ve been afforded access to and coming up with new ideas for it. We’re using AI for garbage cleanup and recycling, all the medical analysis going on for COVID-19 at the moment, and for security purposes such as weapon detection in schools and concert venues. Ultimately, the way we’re finding great uses for AI is really through building connections by having people who have really good ideas and also understand the technology, reach out to someone they might have not reached out to before.
To figure out how to best use AI, we likely have to think about what currently are some of the global challenges that exist. Are there any challenges you see today that maybe are a good fit to be solved by artificial intelligence?
Michael: I had mentioned earlier that fairness and equitable opportunities are really important for broader society. So, I think any challenge that is in the space of trying to make sure that a process itself — for instance, hiring — is really fair. As well, challenges in the food supply chain I think are a really exciting place to do a lot of AI work.
One of the things I’m most excited about is the increased activity in space. So for instance, I always think about how we have all these satellites in space, and only more are going up there. It would be catastrophic if just a small piece of debris or material broke off. It would fundamentally change how our ATMs work and all sorts of other stuff. This then makes me think of the potential for companies like “Space Pirates” who would just be up there collecting junk.
Another instance is AI in sports, which I think is a really untapped area to use artificial intelligence. Think of play calling. I mean, I could only imagine a machine learning algorithm being the best play caller in the world — whether you’re a defensive or offensive coordinator. These are just some spaces that I think AI will make a big difference in over the years to come. Also, maybe a little facetiously, any ideas to make Alexa and Siri better; I think we would all really appreciate that!
What’s exciting is that right now, we have the technology to build the future we want to have. It’s people like current high-school students that are going to define the future we have. I also think that the biggest ideas are sitting with people who may be saying “I’m not a math whiz so I can’t do AI stuff”. To that I’d say: Of course you can. If you have a really great idea, then it’s all yours to build the future you want.
Considering the potential impacts of AI on the “Future of Work”, what advice would you give to students on not only pursuing a career that’s meaningful, but also one that will be shielded by accelerating automation?
Michael: I think that’s a great question, and one that we should be asking especially after seeing the effects of the COVID pandemic. It would really be a tragedy if after all of this, we don’t try to embrace new ways of doing something. Something that I tell my teams all the time while we’re dealing with this is “We don’t want to get back to the way things were, we want to go to the way things should be.” That’s the ultimate goal.
For example, in the context of the “Future of Work”, when we look at people like bartenders, waiters/waitresses, and servers, they all have an incredible ability to interact personally with people and connect on a human level. I think one of the best things people in these positions (along with everyone else) can do to prepare for the future is to involve yourself with communities talking about technology right now. You don’t even have to go to a coding bootcamp or anything like that! There are all sorts of Discord servers, forums on Reddit, and many other channels where people are talking about AI. What you’ll find is that the people who are the perceived “experts” in these channels, their greatest skill set may not necessarily be interpersonal communication; whereas people who are bartenders and servers have those traits and skills. So, go into their communities and bring those traits and skills to them. They’ll definitely be grateful for it.
It’s this complimentary type of effort that I think people could really seize on right now and bring something to these communities that didn’t exist before. Also, Google search is a wonderful thing. Just Google “Open AI Education” and there’s a ton of resources. I’m at MIT right now and there’s a program called, “MIT Open Education”. These platforms help you learn about things relating to AI, and it doesn’t just have to be the hardcore math. That’s one of the reasons I’m inspired by organizations like The AI Education Project that say “Hey, we can focus on different aspects of AI like Humanities or simply being strategic thinkers.”
To end off, you recently published your book “T-Minus AI”. What inspired you to write the book and are there any key takeaways that you think students who read it will benefit from?
Michael: Well, I’ve been in this field for the better part of half a decade now, and I’ve experienced various aspects whether it’s government focused or in the private sector. What I realized is the usual communication barriers that get in our way (with just about anything in life) is that we often speak above, below, or through one another; and we need to get to a place where we can communicate effectively.
Generally speaking, AI is hidden behind many complex technical terms. As mentioned before, it’s also a little bit confused and exacerbated by these extravagant depictions of science fiction — which makes the realities of the technology hard to decipher. However, I do think they’re crucial to understand. AI, in it’s essence, is a really multidisciplinary topic. You kind of need to know a little bit about your own evolution, how we learn, and basic ideas of how a computer works. That’s because we design AI on the basis of our brains, so we should probably know a little bit about our brains as well!
Over the course of these past six or seven years, when I was explaining AI to all walks of life — whether it be at the middle/high school level, politicians, and other leaders of this nation — I realized that there really wasn’t any sort of treatise on the topic that covered the many disciplines mentioned. If you don’t understand how your brain works or how language and writing came to be, then you really can’t discuss AI. These are all really human anecdotal stories, and the way we as humans have learned since the dawn of time is through conveying stories to one another. So, my idea was to write this book to tell lots of human stories to explain AI. Towards the end of the book, I then explore some of our current realities and the state of the world as it pertains to AI.
My goal with the book was really just to bring it home so that everyone was taking away something not only interesting, but that also helps them in life and can serve as a conversation starter. Whether you’re a sociologist, interested in psychiatry, or you really want to be at some place like MIT coding, I hope the book really brings the narrative altogether.
