The Machine Learning Center at Georgia Tech (ML@GT) Speaker Series draws leading researchers from across academia and industry each semester to present on current topics in the field, applications for machine learning and artificial intelligence, and related experiences. The series is wildly popular, with attendance averaging over 100 guests for each event. This spring, ML@GT invited nine experts to campus.
In April, the center hosted Manuela Veloso, head of J.P. Morgan AI Research. Veloso is currently on leave from Carnegie Mellon University where she is the Herbert A. Simon University Professor in the School of Computer Science, and former head of the Machine Learning Department. She is the past president of the Association of the Advancement of Artificial Intelligence (AAAI), and is the co-founder, trustee and past president of RoboCup.
Throughout her career, Veloso has been honored with numerous awards including being a fellow of the ACM, AAAI, IEEE, and AAAS, several best paper awards and the Allen Newell Medal for Excellence in Research.
Her campus talk on “lasting human-AI interaction” filled the Marcus-Nanotechnology Center auditorium.
While she was on campus, we had the chance to chat with Veloso about her experience in industry versus academia, what it means to be a woman in leadership in two male-dominated fields, what she perceives as challenges in AI, and more.
What first drew you to AI and a career in tech?
So, I was actually never enthusiastic or particularly interested in science fiction when I was a kid, but I did like math a lot. I became very interested in automating things and using computing to automate processes, especially after completing my degree in electrical engineering. Getting computers to do more than numerical computations is how I got drawn into artificial intelligence.
How has your experience differed between industry and academia?
I’ve been teaching for more than 30 years, and I’ve only been on the industry side for about nine months, so it has been a big shift coming to work in the corporate world.
One thing that I like about coming to corporate is that there are all of these new challenges. I don’t see problems; I see challenges and good things. I am very excited and in awe of all of the societal problems that I can explore, not just because that’s what I want to research but because it can really impact society.
Of course, the functioning of day-to-day life is very different. There are no students, a completely different hierarchy and organizational structure, so many meetings, and different prioritization of goals. You are no longer just yourself pursuing research with your students. I’ve found the whole transition very exciting.
What are some of the projects that you are currently working on?
At JP Morgan, we are focusing on several different projects. As head of AI research there, I have been working on building my team. I started with people who have a great vision in terms of research and who have experience using AI and machine learning specifically with finance.
Since then, we’ve been adding more people to our team who care about AI to tackle projects related to data. We have data sets with millions and billions of data points in them to sift through. Another concept we care about is simulating environments like multi-agent simulations, explainable AI or making data more transparent. My team also works on data ethics and bias.
What is something that excites you about the future of AI? Any idea on what the next “big/hot” topics are in AI?
I think I have a realistic view of AI. I believe we still have a lot to develop. Meanwhile, a lot of people talk about AI as if it’s done and someone could go to Whole Foods and buy AI. This type of science has made a lot of progress, especially since the introduction of deep learning, but it’s still a very brittle science and is hard to generalize.
There’s AI that helps people read email, but that same AI can’t be used to categorize images or work in finance. We can maybe use the same methods, but it’s not a plug-and-play science. We are making progress in making AI and machine learning more robust, but we are not yet completely done.
What advice would you give to someone going into this field, especially a woman?
My advice to every young person, and especially women, is to not become desperate to be given the credit that you think that you deserve. If you aren’t given credit for your work at some point, keep moving and prove them wrong. Keep working on your research because eventually, if you have the right facts and data, they cannot ignore you.
Find people that you work well with and that have a passion for the science and for the work itself. You cannot be stopped if your work is good.
What do you think is the most exciting part about working in AI?
There is something about AI that is appealing to everyone, and I think it’s especially appealing to women. This concept of having a computer that can help with tasks and provide assistance in making a decision – like which house to buy, how to decorate your room, or which doctor to go to – is very appealing to people. We currently google this kind of information, but I think eventually AI systems will be so “in the loop” with us that they will be able to give personal advice or do things like book a trip to Paris and understand what type of hotel we would like or our seat preference. It’s appealing to think that we can create a machine that we can direct and it does everything for us.
This is also exciting because instead of making robots that play soccer or robots that go to Mars, we are starting to make robots that help with human life. This shift is making AI more transparent and is leading us to ask the question “why” more often.
If an algorithm recommends a book to you and it’s not a good book, that does no harm. If an algorithm decides someone can have a loan versus not having a loan, or someone has cancer versus doesn’t have cancer, then that can cause a lot of damage to someone’s life.
Current machine learning methods that decide which book to recommend are the same as the ones that decide whether or not someone has a tumor, and we need to be more aware of the impact that this algorithm can have on someone’s entire life. I’ve become a bit obsessed with this part of AI and it excites me to work on challenges that have this level of impact.
What is something that you wish you knew at the beginning of your career?
When I started working with robots, I didn’t think about how much of science is a community effort and how important it is to have people who care about what you do and give you feedback. I was very lucky to have a wonderful community, but it is not something that I thought about upfront.
We have this idea that scientific discoveries are made alone because we have these pictures of Madame Curie alone in her lab or Albert Einstein tinkering with something by himself, but that’s not really what science is about. It is great to make a big discovery, but you never do anything alone. It’s important to share what you do, and to listen to and give feedback.
What does being a woman in leadership and in multiple male-dominated fields mean to you? What challenges (expected and unexpected) have you faced?
I’m actually asked this question fairly often and I always find it difficult to answer. For the longest time, I never thought about myself as a woman doing research. I just thought of myself as a researcher. Then at some point you kind of realize that you are a woman after all and that you are not a part of the same groups.
Men tend to bond together over activities like going out to drinks or watching sports together, and that wasn’t something I did and I missed that connection with my colleagues. I never thought it was much of a problem at the time, but over time I did notice this.
I also recognize that as a woman with a [Portuguese] accent, that I would say things and was not always given the credit that I probably deserved. It happened all of the time, and still happens, where I would say something and hear nothing, and then my male colleague would say the same thing and it would be “wow, that’s a great idea!” Sometimes it bothers me, but most of the time I’m just happy that the idea is getting through and going somewhere.
It’s funny, when you’re young you think this is happening because you are young, and then you’re less young and think it’s because of that, but then you realize that you can always give yourself an excuse for why something is happening the way that it is.
I also have to say that in spite of all of this, I think I have been pretty successful and much of that has been due to male colleagues giving me credit. I’m sure that many committees that evaluated me and gave me the honors that I have were not made up of women.
What is a misconception that you think that people have about the tech industry? What’s the reality of it?
Everyone is talking about AI as if it is a solved problem. It is definitely not solved.
They also see all of these big tech companies and the free food and the other perks and think working there is some glamorous life. No, it’s a lot of hard work. Computer programming is very hard and it’s very hard to build upon other problems that have been solved or that others are trying to solve. It’s not a skill that you can pick up in one class or in one book. If you want to be successful, you have to study it and know the field in depth.
Who is someone that inspires you and why?
I am very inspired by my husband. He is a true scientist, an electrical engineer, with a passion for science and for research. I have always looked up to him, and he has helped me a lot over the years. It’s been a great partnership in family and career. We’ve been married for almost 40 years, and he helped make our family possible.
As a sidepiece of advice to other researchers, choose your partner wisely. A career in science and academia is full of ups and downs. It’s important to have someone by your side who gets it and is able to encourage you and to challenge you.
We noticed that you are quite the world traveler. What is your favorite place you have been to and why?
The place I like the most in the world is Portugal because that is where I am from. But my favorite place to travel to is Paris. Paris is the first place I ever traveled to abroad. I was 13 and with my parents, and it will always be a magical place to me. I’ve been there over 100 times and I plan to go many more.