Improving medical technology with artificial intelligence


As a YouTube influencer known for her social media videos, Jordan Harrod translates developments in artificial intelligence and public health to mass audiences.

Harrod, a doctoral candidate in medical engineering and medical physics in the Harvard-MIT Health Sciences and Technology program, is known for her viral videos such as: “Can AI Recognize You From Your Walk?”, “Edge Computing, Explained,” and ” How AI learns to cheat.” When she’s not exploring how humans interact with AI and other technologies, “for better or for worse,” she’s been working on her Olympic lifts at the gym or reading fantasy novels.

YR Media chatted with Harrod on Zoom about what drives her, the future of AI, and the benefits and betrayals of recent AI technology.

Cureha Mitchell, YR Media Contributor: What project have you worked on and what are you most passionate about? How does it work, what was involved in creating it, what makes it special?

Jordan Harrod: My first foray into machine learning was in machine learning and medical imaging. And it was a very interesting learning experience for me in terms of getting into the field.

We want to design machine learning systems that can identify magnetic resonance images (MRI) that have been acquired incorrectly. Because sometimes mistakes made by MRI or software can look like tumors, or something that could be diagnosed as a problem in a patient, when it’s not really there. This project was exciting because it was one of the first projects I worked on that seemed to have a short-term impact.

CM: Can you tell us about a person or an experience that you think set you on the path you are on?

JH: My dad basically told me to ask for what I wanted from a young age because the worst answer you can get is “no”. And it kind of became the motto for most of the decisions and opportunities that I’ve been able to make throughout my educational journey.

CM: Is there a void in your field and how do you see your contributions to filling that need?

JH: At the intersection of my scientific work and my academic work, there is definitely a void of researchers, who are also interested in communicating their work in ways other than giving lectures. And I hope the content work that I do helps fill that gap in people reaching out to audiences that probably wouldn’t encounter traditional forms of public science engagement.

CM: Can you think of a time when you’ve ever considered giving up on your dream? Tell us what you did to overcome this.

JH: I thought I was going to go into the pharmaceutical industry right out of undergrad. I did an internship in pharmacy. I realized that I wasn’t really interested in getting into the pharma industry after all, and I don’t think I gave up on a dream, but it was just a change in the trajectory of the path I was on. was.

CM: What advice would you give to young people looking to break into your world?

JH: The environment you work in is just as important, if not more important, than the work you do, but it’s often not something you necessarily consider when deciding whether to pursue new work, new opportunities, or when you reflect on past opportunities.

CM: We know you can’t see the future, but if you had to predict: what will happen in your field differently than now?

JH: Hopefully we will have a better understanding of the human brain than we do now. I think on the machine learning side, hopefully we’ll start to see more and more cross-disciplinary work, more incorporation of the ethical side of things into the kind of educational pipeline so that people have an understanding and our respect for this side of things.

CM: Do you have any opinions regarding the use of AI in synthetic speech such as Uberduck? Do you see advantages and disadvantages to this technology?

JH: Synthetic speech has some advantages, often in the form of dubbing, which makes it possible to take content originally created in one language and realistically duplicate it in another language. This is also related to deep video fakes where you can make lip sync work better.

At the same time, there are obviously plenty of ways to abuse this stuff. It’s something that’s very easily exploitable, especially since it’s readily available on the internet and you don’t actually need any real expertise to find systems that will let you do it.

CM: How do you want your legacy to be remembered?

JH: I hope that the research I do will contribute to the understanding of the world we live in or to the advancement of human health. But I don’t think I need to be someone people read in textbooks; I just hope the work I do ends up having a net positive impact.

Source link


Comments are closed.