AI or Not

E043 – AI or Not – Sahaj Vaidya and Pamela Isom

• Pamela Isom • Season 2 • Episode 43

Share episode

Welcome to "AI or Not," the podcast where we explore the intersection of digital transformation and real-world wisdom, hosted by the accomplished Pamela Isom. With over 25 years of experience guiding leaders in corporate, public, and private sectors, Pamela, the CEO and Founder of IsAdvice & Consulting LLC, is a veteran in successfully navigating the complex realms of artificial intelligence, innovation, cyber issues, governance, data management, and ethical decision-making.

Artificial intelligence is rapidly transforming healthcare, but with great innovation comes the responsibility to implement these technologies ethically and effectively. Senior Research Associate Sahaj Vaidya takes us on a fascinating journey through the evolving landscape of digital health, offering invaluable insights from her work at Koita Centre for Digital Health at Ashoka University.

Sahaj's path to becoming an AI governance specialist began during the pandemic when she witnessed how automated systems could spread misinformation, inspiring her to champion responsible AI deployment on a population scale. She challenges common misconceptions about digital health, explaining that it extends far beyond digitizing medical records to encompass preparing for future technologies and helping people adapt to AI-enhanced healthcare as a normal part of life.

We explore the tangible benefits already emerging from AI integration in healthcare—from automated transcription services that free clinicians from administrative burdens to telemedicine platforms that optimize scheduling for both patients and providers. The conversation delves into sensitive issues surrounding medical data collection and privacy, including innovative approaches like using AI to create synthetic data for clinical trials without compromising patient confidentiality.

The discussion turns to AI governance, which Sahaj likens to the rules of a football game—essential for preventing chaos and ensuring fair play. She highlights the gap between organizations claiming to have AI governance policies and those implementing them substantively, noting the growing demand for specialized roles like AI ethics officers. The European Union AI Act serves as a case study of regulatory frameworks with global implications, affecting any organization with European customers or operations.

For aspiring professionals in digital health and responsible AI, Sahaj emphasizes understanding core principles before implementation and evaluating whether AI is truly the appropriate solution for each challenge. She envisions healthcare's future as human-machine collaboration rather than replacement, with AI complementing human decision-making in sensitive contexts.

Join us to hear why AI literacy matters for everyone, regardless of technical background, and discover why Sahaj believes AI isn't here to take our jobs but to handle routine tasks while creating new opportunities requiring distinctly human skills. This episode offers essential guidance for navigating the exciting yet complex intersection of healthcare and artificial intelligence.

[00:00] Pamela Isom: This podcast is for informational purposes only. Personal views and opinions expressed by our podcast guests are their own and not legal advice, neither health tax, nor professional nor official statements by their organizations. Guest views may not be those of the host.

[00:51] Pamela Isom: Hello and welcome to AI or not, the podcast, where business leaders from around the globe share wisdom and insights that are needed now to address issues and guide success in your artificial intelligence journey as well as your digital transformation journey at large. I am Pamela Isom and I am your podcast host and we have a unique and special guest with us today, Sahaj Vaidya, and she is a Senior Research Associate at Koita Centre for Digital Health, Ashoka University. Sahaj, welcome to AI Or Not.

[01:34] Sahaj Vaidya: Thank you Pamela for this invitation because it's been a real wonderful time that we had during our brief collaboration when working on the AI governance, responsible AI side of things, working on some, some of those core concepts and things which were, which are kind of very relevant at the moment. So when you reached out on LinkedIn with the invitation for this podcast, I was really more than excited to get on the stage with you and share my thoughts and opinions with the global audience. So this is really a fantastic opportunity and thank you so much for considering me for this.

[02:15] Pamela Isom: Absolutely. I don't forget when I'm collaborating with people and our experiences. So it's a delight to have you here. I'd like to start out by asking you to tell me more about yourself, your career journey. I know you worked with me for a little while, but tell me more, tell me what's in your future.

[02:36] Sahaj Vaidya: Sure. So I was pursuing my PhD at the New Jersey Institute of Technology from 2019 to 202022 and the focus of my research during my doctoral journey was utilizing visual analytics for evidence based communication and gradually then integrating that into responsible machine learning algorithms in order to promote fair and transparency in the communication through charts and graphs. However, there was kind of an inflection point during that journey because a couple of months into my PhD program and the pandemic spread out, so all of us were kind of forced to work from home and we used to scroll through X in order to get, you know, any information that we could gather. And as all of us know that there were a lot of conspiracy theories, misinformation going on around because of the usage of these automated bots and everything. So that was kind of a turning point, I would say, which made me realize that if people can use AI for these kind of things, why not utilize it on a population level scale in a responsible manner so that people can get maximum potential benefit out of it. And there began the journey of my stepping into this domain of AI governance and ethical AI. Currently I'm serving as a Senior Research Associate with Coita center for Digital Health at Ashoka University here in India. And it's been a really good time here in order while working on some of these interdisciplinary concepts of digital health, AI governance, responsible AI. So this multidisciplinary collaboration has also been a very good opportunity for me to leverage some of my past experiences.

[04:41] Pamela Isom: Good for you. Tell me more about digital health. How's that, how's that going? What's that mean in the AI world?

[04:48] Sahaj Vaidya: Sure. So whenever we talk about digital health, what comes to, you know, generally just speaking of digital health, what comes to people's mind is digitization of maybe health records or producing some summaries of patient doctor communication. That is all what comes to our mind. But actually digital health is about projecting the future technologies, preparing for what's about to come with the help of these emerging technologies. AI is obviously the current hype around us, no matter whether it's digital health, fintech, education or anything. But given that in a very sensitive and regulated setting like healthcare, when AI is deployed, it leads to a lot of concerns because of the nature of data and the patients that we deal with. When you tell someone like a patient goes for maybe a very simple diagnosis, and when you tell someone that a robot will perform a diagnostic screening for you, then be kind of little suspicious about revealing their information to that non human entity. So those are kind of some of the barriers that occur when adoption of AI in a commercial setting or when it's deployed on a larger scale. And that is where the digital health comes into picture that we are kind of working towards helping people adopt and adapt to these practices as a norm of life, because AI is here to stay.

[06:25] Pamela Isom: Okay, so then are we advancing as you would expect when it comes to digital health? Or what do you think from a governance perspective? Where are the opportunities?

[06:42] Sahaj Vaidya: That's a very good point because when we talk about healthcare, as I mentioned, that even when AI was not introduced in the healthcare domain, it had a lot of regulations and guidelines to abide by because obviously we deal with human beings or we deal with lives of a lot of people. And now with the integration of AI, it has become challenging. But at the same time the innovation in this area is also blooming every day. You see like there are so many new developments taking place in healthcare sector, even some of the most, you know, basic things like Transcription of the patient doctor communication has become digitized. You don't need to kind of spend, the doctor does not need to spend kind of hours and hours writing down the clinical summaries. So a lot of those mundane tasks, admin kind of tasks are taken away from the doctors so that they can allocate the time to, to their patients. Another trend that's been currently on an increasing graph is the usage of telemedicine. AI and telemedicine because we know that for government hospitals or healthcare centers there's a lot of outdoor patients OPDs. So scheduling appointments and all often becomes very cumbersome in order to reduce the load on the patient side as well as on the side of the clinicians and the doctors. Usage of AI for the scheduler with respect to telemedicine has become a very good trend and that has also led to lot of ease on the part of patients that they don't have to spend hours and hours in the waiting line just looking out for their turn. They, they can just log onto the app and see what's the availability, what's the next next turn and they can adjust their schedule accordingly. Other side for the doctors as well, they know how much time they need to allocate and how many patients they can expect. So it's kind of a win win situation when these kind of tasks are automated with the usage of AI.

[08:52] Pamela Isom: Okay, I like that. So I'll tell you that I'm really interested in digital health and AI in the health care space to help with the diagnosis of situations because our genetic makeup is so unique and so the more data that can be collected, the more the possibilities for cures if we use AI in the proper way. And so I'm really excited about that because I think that we are facing health situations where we need all the help we can get. Right. So but what we need is accuracy. What we need is accuracy. So that's where governance comes in as one of your concerns, one of the things you've alluded to. But we sure need the help.

[09:44] Sahaj Vaidya: Yes. And you actually made a very good point when you said about collection of data. That's also one of the major things with respect to healthcare because a lot of patient data is kind of sensitive in nature. It's like their past medical records, their medical history, the current symptoms they're experiencing. So even though we have a couple of open source data sets, but when you want to train your AI models or when you want to look at some new emerging trends, collection of medical or health records is Very challenging, especially with an ethical aspect like, you know, there are a lot of regulations with respect to the informed consent getting the users on board. Or there's also this new concept of, you know, utilization of AI for clinical trials, for dummy clinical trials. Because you know that generally when users or participants are recruited for clinical trials, a lot of conflicts and concerns occur because they have to share some information, then you get back to them, some experiments are performed on them. So it kind of raises tons of questions for the safety and security of the patients as well as their data. So new trend these days is emerged to utilize AI for creating these kind of dummy trials. Like you create dummy participants to simulate the real world scenario and that's how you go ahead with your study. So in this way you won't have to bother recruiting people and you won't have to go into the details of these informed consents and everything and your task is pretty much solved. But again, utilization of synthetic data and the corresponding accuracy is also a question. The accuracy, utilization of data trade off plays a major role here.

[11:39] Pamela Isom: I like your passion around doing something to benefit society. Right? Use it and do something with it. As going back to what you started out with in your initial comments, I do have a question for you. First of all, congratulations on your Elle India award. So congratulations on that. And I'd like you to tell me more about that. What was involved in that award? Why did you get it?

[12:09] Sahaj Vaidya: So actually a couple of months back, this LA India magazine had reached out to me through some of the online platforms asking about my work, my previous research, what I was doing at the moment. So it was kind of a casual conversation about all that I had been doing. But I really did not expect to be selected because there's also this thing that we see people going into this AI domain working in these fields. But as you might have noticed, and you might know of this more than I do, there's very less representation of women in these fields, even though these are emerging technologies. And we kind of see a lot of skewed representation at the moment. But I do believe that since we are already talking about this, I want to kind of convey message to this audience on a global level that AI is not something that you need to learn, or you, if you want to learn, you need a very technical computer science, data science background. No, AI is something which maybe anyone can learn irrespective of their prior background and experiences. So if you feel you want to just jump into this field, just jump in. There are lots of opportunities to learn, to apply and gradually move ahead in this domain. So just don't hesitate. If you like this domain, go ahead. The world is full of opportunities.

[13:43] Pamela Isom: Well, I would say, again, congratulations on your award and I think you're being very humble here. I know some of the work that you've done, so I know that you are very deserving. So that's a very good thing. And the fact that we think the fact that you have passion around AI literacy, I want you to talk some more about that. So from an AI literacy perspective, which you say we should all feel and know that there is a place for us when it comes to AI, particularly women, because you're seeing more men in the field than women, we should all feel that there is space and room for us, which to me touches on AI literacy. But I'd like to give more of your perspectives on what you mean by AI literacy. And again, where are those opportunities? What are we doing well, and where are those opportunities?

[14:40] Sahaj Vaidya: So actually, right now, because AI is a very rapidly changing domain, there's a lot of kind of hype surrounding it, and people also kind of do not have a very clear understanding of what it is and what the entire landscape looks like. And that's why I believe in order for people to adopt the utilization of AI easily, first it's necessary to instill that basic foundational level of AI literacy, not just amongst the schools or, or the university students, but for all professionals or for all people at all kind of level, irrespective of their background, should be imparted this basic level of education for AI. Otherwise, what may happen is there'll be a lot of misinformation going around that AI can do this, AI cannot do that, and people will fall prey to those conspiracy theories. But in order to dispel those doubts, it's necessary that people understand what is AI and what it can do for them if used in a responsible and ethical manner. And here I would also like to kind of digress a little bit towards the EU AI act, which has kind of mandated the literacy of AI at certain levels. And as you know, that just on 2nd of August 2020 25, the General Purpose AI Code of Practice came into effect in the European Union. I believe that's a very good step towards governing AI systems, be it high risk, low risk, or a general purpose. But I believe it's a good step in the direction of ethical AI and AI governance at a population scale.

[16:32] Pamela Isom: Yeah, no, I think those are good, good insights there. I appreciate you bringing up the act. Right. The European AI Act I don't think a lot of folks have, have digested it yet. And I still think that what's going on in society today is if in your, if you're in the us you don't think that the European AI act is applicable to you, but honestly it has a global reach. And so I'm glad that you brought that up here because it matters to all of us.

[17:03] Sahaj Vaidya: Exactly. Yes. And it has far and wide implications. Like if you have, irrespective of whether you are based in EU or not, as long as you have your customer base or maybe your offices or employees within the European Union region, you have to abide by those compliances. And that's why that level of literacy becomes of necessary importance.

[17:26] Pamela Isom: Okay, so you think one area of literacy should be along the lines of understanding some of the key regulations like the European AI act, and really start to get a good sense of some of the policies associated with it because it has global implications.

[17:44] Sahaj Vaidya: Yes.

[17:45] Pamela Isom: Okay. And plus, I just think it's just basic foundational knowledge that you're, you're seeing this, in my opinion. But I think I hear, I think I heard you say through all of that, the European AI act, yes, it is a somewhat of a regulatory compliance thing, but it's so much more than that because it helps instill some of the AI fundamentals and some of the literacy components that are needed as far as how AI could be applicable to you and how to use it responsibly. Is that correct?

[18:18] Sahaj Vaidya: That's absolutely correct. Because generally, even right now, even though the act has come into effect and will go into full implementation next term, there's still a lot of confusion surrounding what companies need to do or what people need to do, what they need to abide by, what kind of certifications are needed. There's no very clear cut guidance at the moment.

[18:43] Pamela Isom: Yeah, yeah.

[18:44] Sahaj Vaidya: And that's why I think spreading that awareness becomes significant.

[18:50] Pamela Isom: Yeah, No, I appreciate you. I was reading and I know this to be true, but 77% of organizations are already working on AI governance and it's expected to rise because the perception is that AI there's concerns around AI sprawl. Enterprises are challenged with things like AI sprawl because they're in a hurry to just implement in order to meet certain public perspectives. There's concern about a need for interoperable infrastructure and we need governance in order to help fuel and drive that. We don't necessarily want centralized governance, but we do feel like that to some extent that would be helpful in order to optimize efficiencies across organizations so that things are not so siloed and adoption is not as haphazard. So I think of AI sprawl. I'm very particular about interoperable infrastructure and paying very close attention to the infrastructure as we are looking at integrating AI within our environments and also as we are building AI specific products. So my perspective on that is AI governance is critical. In order to address some of those challenges, I would like to give your perspective a little bit more on AI governance and how do you see it adding value?

[20:30] Sahaj Vaidya: Sure, you really made good point there. Which reminds me of an analogy that this AI governance is like the rules of a football game. If, say, for example, there were no rules or if players did not follow the rules during the game, then what happens, what the consequences are? They may be removed for the game, they may be banned for a couple of years, they may be suspended. A lot of different things could happen. Similar is the case here. If we let the AI systems run wild without putting any safety guardrails surrounding the usage of these systems, then we know that we cannot prevent the mishaps or the untoward accidents that arise from it. So that's kind of a very simple analogy to explain to people the importance of AI governance. As you mentioned that almost 77% of companies currently are working towards, you know, achieving the goal of utilize of governing their AI systems. But often what happens is that they term these things like we have AI governance, we have AI policies, a lot of those things. But when you actually look into it, maybe they are not so proactive in their approach towards the governance. They used it maybe because it was mandated by their authorities or something like that. But I've often seen here as well as on a global level that those are just namesakes. They have maybe two, three documents outlining some policies and some bullet points here and there. But actually when you ask them like, hey, what are you doing for, you know, governing your high risk system since you are dealing with healthcare, they would not have a very clear cut answer. And that's why I think that with this emerging landscape, the role of AI Governance Analyst Chief AI Ethics Officers becomes all the more important responsible. We are already seeing a lot of job openings for the same like responsible AI Officer, Chief AI Officers, Ethics Chief AI Ethics Officers A lot of those jobs opportunities are popping up. So I think that is a good direction if the people can really understand the essence behind it and what requires to be done.

[22:57] Pamela Isom: Yeah, I agree. So what advice would you give to early career professionals like yourself that are interested in digital health research or policy.

[23:08] Sahaj Vaidya: Actually this is kind of, I would say a very interdisciplinary thing. Like as I mentioned that I have a kind of earlier I had a computer science background, then I moved to data science, visual analytics and gradually entering this domain of responsible and ethical AI. So I would say that if you are interested in digital health and responsible AI space, one of the best thing to do here is to kind of first have a very proper understanding of what are some of the core principles of responsible AI and what is AI governance. Because you see a lot of research literature surrounding AI governance and things. But as you might have realized that there is no single concrete definition of what is AI governance. And that's why I believe that diving deeper into this domain in order to establish some golden standard for the same might be very helpful. So first get a very good sense of what the literature looks like, what are some of the existing gaps or some of the niche problems that need to be addressed. Some of these low hanging fruits. Often we think of AI solutions for problems that might not even exist. But just because AI is the buzzword, we try to put it into every single thing that we come across. But not every problem requires an AI based solution. It could be solved by something even simpler and maybe at a fraction of cost. So just think of it before you try to attempt or solve any problem that whether it really requires an AI based solution. If yes, go ahead, but make sure that you put safety card, make sure that you have an understanding of what you are about to implement.

[25:01] Pamela Isom: Okay, yeah, that makes a lot of sense. So tell me more about your transformative actions and outcomes. So have there been any moments in your work that were particularly transformative or surprising?

[25:24] Sahaj Vaidya: So almost two to two and a half years I have been kind of taking baby steps into this domain of AI governance and ethical AI. And some of the best moments that I could recollect of was my first hands on experience with responsible AI was when I started working with you over February 2024. And I actually got to learn that how those policy documents would look like in a real world setting. Because earlier I just seen through some of the LinkedIn posts or written papers on it. But we of course don't want them to stay in a paper, we want to bring them to real world settings. And I think that my journey for the actual, you know, leveraging my past research began and I started learning how to practically deploy these things, how to integrate them with some of the existing AI systems. So I think that even though that might be a very Brief collaboration, but that really made a major impact on some of the work that I did after that.

[26:32] Pamela Isom: Well, let me tell you, I appreciate that. That's what I try to do when I onboard early career researchers and things. I want to give them an opportunity to get real life practical experience because there's nothing more significant than when you can take your research and put it into practice. So you should know that the policies that you created are in use today by the organization. So the policies that you helped pull together definitely know that they are in effect, they are in use today and continuously evolving. So thank you for what you did and I have you in mind for the future. And so we will definitely make sure that we stay in touch. But I'm glad that you brought that. I didn't know you were going to do that, but I'm glad that you did. And I could see that as being a pivotal point and step in your career and I'm just honored to have had you work with me.

[27:26] Sahaj Vaidya: Yeah, I really appreciate that.

[27:29] Pamela Isom: Absolutely. So, that being said, I don't know if there's anything else you wanted to share about digital health or anything, so if you do. I had been thinking about regulations, but you touched on the European AI act, so I think that that was plenty. I don't know if there's any others that you want us to pay attention to, but I do think that that was very important. So I'll ask you, is there anything else that you want to share with me and the listeners before I ask you your final question, which is can you share words of wisdom or a call to action or both?

[28:03] Sahaj Vaidya: Yeah. So as far as digital health is concerned, I think there are a lot of different pathways emerging here. People have been kind of working on integration of AI into existing systems or, you know, introducing something very novel like robotics surgeries, which are already taking place at a couple of places around the world, especially in the European Union and some of the places in us. So those are very transformative applications of AI that have been seeing lately in the space of digital health. And I think that marks a good progress towards, you know, taking a step towards actually not automating things, but in the direction of the human machine collaboration that I've always envisioned, especially in sensitive setting like healthcare. In this setting, you cannot expect AI to completely automate the decisions or take its own decisions because you are dealing with life of people. So you need human in the loop kind of approach here where the AI complements the human cognition, the human decision, so it helps them make decision. But not replace it.

[29:19] Pamela Isom: Okay, all right, that's awesome. So now tell me either words of wisdom or a call to action or both. What you got?

[29:29] Sahaj Vaidya: Oh, so that reminds me that often when I scroll through LinkedIn, I see a lot of people, you know, mention that after maybe say 2030, that's about like five years from now, X percent of jobs will be replaced by AIs. Robots will be doing our jobs. And that spreads a lot of apprehension among people because obviously for someone who does not have a proper understanding of what AI is and what it does, they of course feel like, oh my God, they will feel insecure about their jobs. But let me make clear picture here that AI is not here to replace our jobs. It is here to kind of complement what we have been doing, to take that mundane task that, you know, those repetitive tasks that do not require the human level cognition. In fact, AI has been creating new job opportunities, as you might have seen lately, like prompt engineering, AI engineers, AI officers, a lot of those opportunities have popped up. And of course those cannot be done by AI whenever there is human level thinking, intellect involved. Those tasks can never be done by AI. And I think that's why it's called artificial intelligence. And it will always, I envision it as working in a human machine teaming kind of approach rather than one replacing the other. So don't be afraid. AI is not here to take away our jobs. It's just there to make it easy, make it simple. So don't be afraid of it.

[31:10] Pamela Isom: I love this discussion. Well, I really want to thank you for joining me today. And thanks for those words of wisdom and that integrated call to action which is embrace the capabilities, don't run from it, and don't be afraid of it. So I appreciate that. I appreciate you being here today day and we've had a great conversation.