Anne Kim Podcast Transcript
Anne Kim joins host Brian Thomas on The Digital Executive Podcast.
Welcome to Coruzant Technologies, Home of The Digital Executive podcast.
Brian Thomas: Welcome to The Digital Executive. Today’s guest is Ann Kim. Ann Kim is co-founder and CEO of Array Insights, which is based on her graduate work at MIT. She brings a wide range of technical knowledge and business experience to the venture.
And work with Professor Alex Sandy Pentland, Principal Investigator of the Human Dynamics Group at the MIT Media Lab on federated learning and blockchain solutions for clinical trial optimization using open algorithms or opal outside of her research and has worked on a number of different projects in computer science and technology.
And molecular biology, such as quantum chemistry simulations, genome wide association studies, natural language processing for electronic health records, machine learning and cyber biosecurity work with the EFF and DEF CON. Well, good afternoon and welcome to the show!
Anne Kim: Thanks!
Brian Thomas: Appreciate you jumping on. And I know Sometimes traversing time zones, you’re out of New York City.
I appreciate that staying up a little bit later this evening. Again, love jumping on podcasts. No matter the time of day. So, and let’s jump right into your 1st question here. Could you share with us what inspired you to pursue a career intersecting technology, biology and health care and how your work at MIT with Professor Alex Sandy Pentland propelled you forward towards founding array insights?
Anne Kim: That’s a great first question, Brian. So admittedly, I didn’t want to be a founder. I had no inclinations towards that. I just wanted to be in research. I saw that the intersection between biology and computational and like numerical methods as a really exciting opportunity to make a really big impact in healthcare ultimately, what I found after a lot of like academic work was that.
It was less about the algorithms and machine learning techniques. I was developing and more about the systemic ways that we were getting access to more diverse data sets. And so, I think from a implementation standpoint, it made a lot of sense to shift from academia to commercialization of technology, namely in privacy, preservation and like, privacy, enhancing techniques.
So that’s when I went to the MIT Media Lab, where there are a lot of companies that are spun out especially out of Sandy Pentland’s group. And so, with him, we created this architecture called Open Algorithms or OPAL, which is an implementation of federated learning that allows you to share Health data without necessarily exposing any private information.
And so, what we’re doing at array insights is using that foundation in order to bring cures without the compromise of privacy or patients and namely with patient advocacy organizations.
Brian Thomas: Thank you. Appreciate that. Love how again, sometimes we’ve got professors in there pushing people, not just on the academia side, but also if they see a spark in the student’s eye, and they should pursue something in your case, entrepreneurship.
I think that’s just amazing. I love the story there. So, and can you explain the core mission of array insights and how your graduate work at MIT influence the inspiration and creation of your work?
Anne Kim: Yeah, so the core mission of array insights is to work with patient advocacy organizations in order to engage patients without compromising their privacy.
I’d say personalization without being creepy. And so, one of the ways you are able to do that is by. Establishing trust patient advocacy organizations are extremely good at this. I think. Out of all the different stakeholders and healthcare, whether you’re talking about hospitals. Physicians, even or certainly insurance companies, these patient advocacy organizations have the most alignment with the patients themselves.
And so that means that they are a really trusted resource for patients. We want to honor and respect that. And so, what we’re able to do is. Connect those patient advocacy organizations to patients themselves to answer a question about how that connects with my work at MIT. Well, we can think of privacy and security as a technical way of establishing trust.
Right, we trust the technology of our phone and trust that we are the only ones able to access it because of the privacy technology that’s behind it. And so that is. Those are the 2 sides, right? Trust from a emotional standpoint and trust from a technical standpoint in like, privacy technology.
Brian Thomas: Thank you. And some of the stuff has really gotten advanced. I know just before we hit record, we talked about a term, but cyber biosecurity. And I think you get things are ever so evolving and I know that’s kind of a broad term, but that covers some of the. Privacy we’re talking about, so I appreciate you sharing that those insights with us to this evening.
And, and you’ve worked extensively with federated learning and blockchain solutions for clinical trial optimization for our listeners who may not be familiar. Could you break down how these technologies work and why are they revolutionary for health care?
Anne Kim: Yeah, so I would say blockchain federated learning.
As well, as other privacy technologies, like fully homomorphic encryption or multi-party computation are under this large umbrella of different privacy preserving techniques that have different facets to them that apply to different applications. And so, with blockchain, you have a trustless peer to peer system where from peer to peer, you’re able to share knowledge with an immutable leisure that the whole community acknowledges.
Is un-tamperable, whereas in federated learning, it can be done in trustless. Architecture, but at the same time, it can also allow for facilitators or like a trusted node in order to do that distributed analysis. So, it’s a little more flexible. And I think in terms of implementation, it’s less. Rigid and much more palatable, I think, for integration with existing systems that we have today.
And so, thinking about array insights we transition from blockchain to federated learning because I don’t think established stakeholders in health care are necessarily ready for blockchain. I think a good intermediary was federated learning when we were pursuing that.
Brian Thomas: Thank you. I, I appreciate that. And yes, we still experience people, you know, a little bit hesitant to adopt blockchain or they don’t quite understand. And it’s been out for gosh, probably 14, 15 years now. And obviously longer.
Anne Kim: Yeah. Longer than the 2008 and the financial crisis.
Brian Thomas: Yep. Absolutely. So, yeah, it’s, it’s, it’s amazing. In fact, our publication here is not only Web2, but we have a Web3 component as well.
And we love to promote. Blockchain technology, so appreciate the share. I really do. And the last question of the evening with your vision of the future. How do you see artificial intelligence and machine learning further transforming health care in the next decade, especially in areas like patient privacy and treatment personalization?
Anne Kim: Yeah, great question. I think with any sort of technology, the ultimate vision is that these integrations become seamless. And so, with artificial intelligence and privacy preserving technology, I hope it’s something that we don’t even need to acknowledge in the same way. We think about like TLS or. Like, those are technologies that make our Internet connections secure and safe or HTML as well.
Or rather we don’t need to acknowledge those systems explicitly, but they’re just kind of on the back end. And it’s pretty easy to just. Tack it on or integrate it into whatever you’re implementing. And so, I hope that personalization, AI and privacy and healthcare will be more of a given and not a feature necessarily.
Brian Thomas: Absolutely. And I appreciate your insights on that. You know, it’s more than just that connection. I think there’s different layers of security that can be applied as well for patient privacy. For example, it’s not just something that you said, something that you got to go do. It’s something that.
Should be part of our everyday lives when we’re trying to protect information or if we’re trying to you know, do some customization around the treatment personalization. So, I appreciate that and, and it was such a pleasure having you on this evening and I look forward to speaking with you real soon.
Anne Kim: Thanks, Brian. The pleasure to you and I hope you have a wonderful rest of your evening. Thanks.
Brian Thomas: Bye for now.
Anne Kim Podcast Transcript. Listen to the audio on the guest’s podcast page.
