The diverse applications of artificial intelligence are transforming industries at an unprecedented pace, but with great breakthroughs come complex challenges.
Today I'm joined by Rajvir Madan, Chief Digital and Technology Officer at Arcutis Biotherapeutics, to break down the transformative impact AI is having in our field.
Our discussion spans from the reduction of drug discovery timelines to addressing ethical concerns related to data biases. We also confront challenges tied to data quality and the implications for the workforce in the era of AI.
To grasp the full spectrum of AI's influence on biotech and the vision it's charting for our future, tune in for a compelling episode.
Specifically, this episode highlights the following themes:
Links from this episode:
The diverse applications of artificial intelligence are transforming industries at an unprecedented pace, but with great breakthroughs come complex challenges.
Today I'm joined by Rajvir Madan, Chief Digital and Technology Officer at Arcutis Biotherapeutics, to break down the transformative impact AI is having in our field.
Our discussion spans from the reduction of drug discovery timelines to addressing ethical concerns related to data biases. We also confront challenges tied to data quality and the implications for the workforce in the era of AI.
To grasp the full spectrum of AI's influence on biotech and the vision it's charting for our future, tune in for a compelling episode.
Specifically, this episode highlights the following themes:
Links from this episode:
Rajvir Madan [00:00:00]:
I see a lot of sort of trends around automation. I think if we look across any organization, there's a ton of inefficient processes, and you're starting to really see the technology sort of get there. I had previously sort of seen technology that helped you automate inefficient processes, but now there's technologies that are even helping you identify what those inefficient processes are and then help you sort of automate and optimize those processes.
Steve Swan [00:00:29]:
Hello. Welcome to Biotech Bytes, where we speak with IT leaders in the biotech industry about current technology trends and how they're affecting our industry. I'm your host, Steve Swan, and I've got the pleasure today to be joined by Raj Madden from our Arcutis Biotherapeutics, where he's the chief digital and technology officer. Raj, thanks for joining us.
Rajvir Madan [00:00:49]:
Hey, Steve, good to be back with you and happy to be on the show. Thanks for including me today.
Steve Swan [00:00:56]:
Nah, you're welcome. And there's a ton of things going on in our industry and several things I want to hit with you today, but I think the big one that we speak about, you and I have spoken about it before, and a lot of folks speak about it, is AI. What are your thoughts and feelings around AI and where it is, where it's going and what's happening with it in our industry?
Rajvir Madan [00:01:17]:
Yeah.
Rajvir Madan [00:01:18]:
So I would say I'm cautiously optimistic about AI. That's how I sort of think about it. I think on one hand, the impact that AI continues to have in our industry is just phenomenal. I mean, think about the use of AI for things like drug discovery and drug development, which ultimately can help reduce cycle times by about 70% to 90% in terms of which drugs actually make it to commercialization. So it's not only helping with the speed of how drugs can get to commercialization, because you're sending more and more better drug candidates sort of through the pipeline, but it's also helping you identify which patients are actually going to get the best outcomes from those drugs.
Rajvir Madan [00:02:07]:
Right.
Rajvir Madan [00:02:08]:
So I think that's just one sort of use case, and I'm sure we'll talk about many others as we go on. So I continue to be sort of really positive about how AI can benefit our industry and has already benefited our industry. I mean, who wouldn't want to be in an industry where you could use AI and you could launch drugs and you could reduce the risk of getting those drugs through the pipeline by 70% to 90%, which improves patient outcomes at the end of the day. Right. But I think for me, why I'm cautious about it and I'm not sort of exuberant about it is because I think there's still many things that need to be worked out with AI. I think there's challenges around data biases. Let's say that we feed some of these models. There's challenges around the ethical use of AI, for instance, that we really have to think about, because in a lot of cases, AI models have been and can be producing misinformation, which is never good for any of us.
Rajvir Madan [00:03:08]:
And then we also have to sort of think about this notion of how do we build data quality before we even get into sort of building some of these AI models? Let's say there's a whole conversation around talent that we need to have, which is what talent do we need around getting organizations ready for AI? But also, how is AI going to disrupt the workforce going forward?
Rajvir Madan [00:03:35]:
Right.
Rajvir Madan [00:03:35]:
There's a lot of talk about job displacement and job replacement and job evolution, let's say. So those are conversations that we got to have. So I think I'm cautiously optimistic. I think it's going to have a positive impact to several organizations out there, whether in my industry or in other industries. But I do think we also got to think through some of the risk factors with AI as well.
Steve Swan [00:03:59]:
Well, there's a lot of them and a lot there to unpack of what you just brought up there. Right. First and foremost, what you hit on the research side of things, right. Just talking big picture. Anytime you go and you see they do these hearings on Capitol Hill, right, where they're talking to the big drug companies, and what are the ceos saying? They're saying drug discovery, time and investment. Well, to your point, if we can cut that, even if we cut it by 10%, Raj, right. We're talking years and we're talking tens or potentially hundreds of millions of dollars. I mean, if you're talking 70, 80, 90%, boy, I mean, imagine that.
Steve Swan [00:04:38]:
That's more than a game changer, right, for the research field.
Rajvir Madan [00:04:41]:
Yeah.
Rajvir Madan [00:04:42]:
And I think what it also does more importantly is the cycle times are important. I agree with you on that. But with the use of AI, you can identify patients that are going to have the best impact on that drug. Right. We can just talk about simple examples like Tylenol and Advil.
Rajvir Madan [00:05:02]:
Right?
Rajvir Madan [00:05:02]:
So Tylenol works for me, Advil doesn't work for me.
Rajvir Madan [00:05:07]:
Right.
Rajvir Madan [00:05:07]:
I'm sure it's the opposite for you, or maybe it's the same for you, but some drugs work for certain individuals and they don't work for other patient. And, you know, discovering drugs with this intelligence in mind that the drug is going to work for someone like Steve or someone like Raj, but it might not work for this population. I think that's really important as well.
Rajvir Madan [00:05:28]:
Right.
Rajvir Madan [00:05:29]:
Because then we don't have to go through this cycle of sort of experimenting with drugs, whether it's in a clinical stage or in a commercial stage, before we really know whether that drug is going to have a positive outcome or not.
Steve Swan [00:05:41]:
Right. And then knowing whether that's. Again, now, I'm going back to some of the things you said initially, taking what you just said, and then being certain that the data you're seeing is real data, it's not synthetic, it's not hallucinated data, it's not average aggregate data that AI has been known to come up with. We got to make sure that we got the right data that's coming in for tweaking our models. One of the other CIOs that I spoke with who's at a large biotech firm who I've worked with for a long time, I asked him, he started when his company was small, and it's a big company now. And I said, what piece of advice would you have for folks at small biotechs? He said, get your data ready. We got 30 years of data. We know that's our data, so we know that's good data.
Steve Swan [00:06:27]:
We just got to get it ready for these models. When we buy data, we still got to go through checking to make sure that that's good data. So get your data ready. Was the advice there?
Rajvir Madan [00:06:37]:
Yeah. And I think, Steve, to that point, it's get your data ready, get your data curated, cleansed of high quality. But I think it's also making sure that your data is representative enough of the population that's out there.
Rajvir Madan [00:06:54]:
Right.
Rajvir Madan [00:06:54]:
You don't want to be building these models with certain segments of the population that are underrepresented, let's say, not being part of the training data sets.
Rajvir Madan [00:07:03]:
Right.
Rajvir Madan [00:07:03]:
So I think that's the other sort of consideration that I often think about when it comes to some of these data sets. And then I think, for me, I think the other sort of interesting thing has been around. What new? I think you're going to start to see the emergence, and you've already started to see the emergence of sort of these adjacent companies that are really starting to monetize some of the data that they already have.
Rajvir Madan [00:07:29]:
Right.
Rajvir Madan [00:07:30]:
I mean, think of a company like 23 andme and the partnership it has with GSK, all of this is public information, by the way, but they've been able to sort of take some of their genetic data, and GSK has been able to use that for some of their drug discovery, drug development, modeling.
Rajvir Madan [00:07:47]:
Right.
Rajvir Madan [00:07:48]:
And I think you're going to start to see more and more companies that provide some of this genetic data, some of this biomarker data, some of this DNA sequence data, let's say, that can be used to evolve some of these AI models. So I think not only is it about data quality and about making sure that you curate your data and you have a representative sampling of your data as you train some of your models, but I think it's also about making sure that you think about sort of the universe of data that's available to you and how those can be used to sort of fine tune your models. And I'm foreseeing that there's going to be more and more of these sort of companies that start to monetize some of the data sets that they already own. I mean, I would love for companies like Labcorp or quest Diagnostics. I mean, they've been collecting blood samples from you and me for many years. Right? Wouldn't we love to sort of feed that in an anonymized way, in a deidentified way, let's say, into some of these models so that we can get more precise about some of the medicine that we discover?
Steve Swan [00:09:01]:
On the retail side, they call it zero party data, right? I don't know. What they call it in research is the same thing.
Rajvir Madan [00:09:06]:
Yeah.
Steve Swan [00:09:06]:
I mean, that'd be awesome. That's a great idea.
Rajvir Madan [00:09:08]:
I love that. Cool.
Steve Swan [00:09:10]:
Well, so I've been talking to other folks, right? Doing podcasts, and some of the pain point has been for the smaller companies. Listen, I go to the big data vendors. We all know who they are, right? I need to get 10 million people's worth of data, and I'm spending millions of dollars. I can't do that. I'm a small company. So one guy made a shout out to data companies. I've already hooked him up with two small data companies because he's like, I need 10,000 people's worth of data, not 10 million. So he's already connected.
Steve Swan [00:09:40]:
He's talking to two other small data companies. So to your point, it's got to start evolving here, right? And I think that's a great. I didn't even think. I don't know, maybe others are thinking about the lab cores and so on and so forth. But I think that's awesome.
Rajvir Madan [00:09:53]:
Yeah.
Rajvir Madan [00:09:53]:
And I think, Steve, I think the other thing I'm starting to see in this space know we have access to data sets that we don't fully sort of squeeze the juice out of completely. And I'll give you an example. Most sort of companies, even small to mid sized companies, have access to EHR and EMR data sets, right? I think most sort of pharma and biotech companies have access to these data sets. But there are doctors notes. When you go into a doctor's office, the doctor sort of writes notes and creates a chart and all of that, let's say a lot of it goes into your ehremr system, let's say. But traditionally, because it's been sort of unstructured data, I mean, it's someone's dictation, someone's notes, whatever it is, a lot of companies haven't sort of completely mined that data. They haven't completely sort of used that data set, let's say. But it's there.
Rajvir Madan [00:10:52]:
It's something that most of us already pay for. Right? So I think if you're sort of a small to mid sized company and you have access to ehremr data, and you can apply things like natural language processing and really mine some of the doctor's notes, the clinical notes that exist in those data sets, and that could help you with things like safety signals, it could help you understand whether your drug is working, whether it has the right efficacy, whether it has any issues with toxicity or safety, for instance.
Rajvir Madan [00:11:27]:
Right.
Rajvir Madan [00:11:28]:
So I would say, even before you go out and sort of buy new data sets, I think, think about sort of what data sets do you already have sort of access to, and which of those data sets do you want to completely sort of squeeze the juice out of? And I think not enough sort of companies are doing that. I would say.
Steve Swan [00:11:47]:
Yeah, no, that makes a lot of sense. Like you said, utilizing what you got and then getting more out of it. I've spoken to one company that they've got several use cases right now for their AI, and one of them was really, like you said, natural language processing, having their scientists go in and show me the data, show me some of the experiments, but you're taking to the next level. I mean, you're really getting more granular than just the experiments and what's gone on in the lab. You're getting down to the individual patient data, which is. That's really cool stuff. That'd be great. So now let me ask you a question.
Steve Swan [00:12:27]:
I haven't really gone this route with other folks. You and I kind of touched on this in the past on one of our phone calls. The perception, the push internally at all these companies for AI, it's coming from a lot of directions. It's in the press.
Rajvir Madan [00:12:42]:
Right?
Steve Swan [00:12:42]:
What are your thoughts around that? What do you think as far as an IT leader, right. A digital leader inside of biotech? There's a lot of pressure for us right now, right, to make this thing, to monetize this thing, to help us to put some ROI on.
Rajvir Madan [00:13:01]:
Think, Steve. I think, look, I started off as an engineer, and I still claim to be an engineer. I went to school for engineering, and I think one of the things, and I can say this because I'm an engineer, one of the things that engineers love to do is sort of find a problem to a solution that they've already created, right? And I think a little bit of that maybe a lot of that is sort of happening with AI as well, right? There's been a lot of really creative, smart solutions that are already out there, and now individuals are sort of looking for problems that those solutions can solve, right? That, for me is the complete opposite way and the way we should be sort of thinking about this, right? As Einstein once said, and I famously sort of quote him, he once sort of famously said, he said, look, if I had 60 minutes to solve a problem, I would spend 55 minutes thinking about the problem and sort of five minutes actually solving the problem, right? But as engineers, we tend to sort of do it exactly the other way. And I think that what that sort of gets me to is this notion of, I don't get too enamored with the technology first. For me, it's really about spending the time up front to really understand what is the value that AI can deliver to my business, to my shareholders, to my stakeholders, to my patients, to my doctors, to my customers. That's the question that I'm asking myself, right? And if I really believe that there is value in certain use cases that AI can deliver, it's promised through, then those are the use cases that I'm going to go after, right? That's the way I have been sort of thinking about AI. I'm not one of those sort of individuals that's saying, let's just go launch 100 pilots and see which ones sort of work and which ones don't work. I think I'm one of those individuals that wants to be really focused in terms of where I invest in AI.
Rajvir Madan [00:15:18]:
Just like any organization, there is limited capacity, there is limited bandwidth, there is limited funding. There is limited capital.
Rajvir Madan [00:15:29]:
Right.
Rajvir Madan [00:15:29]:
And I think being smart about how you deploy those limited resources, whether human or whether capital resources to me starts with a better definition of what the problem statement is and really articulating that problem statement up front, doing a little bit of sort of primary or secondary research up front to really understand whether that problem statement has any legs, what value can it deliver? And then sort of thinking of technology to actually address that problem statement. And believe me when I say this, I have worked on many problem statements that once the problem statement gets defined, you might not even need AI because the problem is a process problem or it's a resource problem or it's some kind of another problem that technology isn't even going to be able to solve.
Rajvir Madan [00:16:23]:
Right.
Rajvir Madan [00:16:24]:
I mean, I have a great example from a few years ago where I was working with this business unit, and they sort of came to me and they said, well, we need to go buy this sort of piece of technology. And when we sort of spent a good amount of time really understanding what the problem statement was, the problem actually could not be fixed with technology. If we would have gone and purchased the technology, it would have been a complete waste because the problem was really sort of a notification problem, like, how do you engage patients, let's say. And we actually did not need technology to do that. So I think starting with a problem statement, just make sure that you're spending your time, your energy and your resources in the right way.
Rajvir Madan [00:17:08]:
Yeah.
Steve Swan [00:17:08]:
And like anything else, like you just said, with anything in it, make sure, like you said, you know what you're going after, what you already have in house, and maybe AI will solve it, maybe it won't. But as long as we understand what we're going after, let's go for it then.
Rajvir Madan [00:17:25]:
Sure.
Steve Swan [00:17:26]:
I think that it seems to me anyway, that there's a lot going on, a lot of push and a lot of talk around AI, and it's brandy new. We're learning how to crawl here.
Rajvir Madan [00:17:38]:
Exactly.
Steve Swan [00:17:38]:
We're not even walking. We don't even know yet really what it's going to. I mean, it's going to be doing something different at the end of this calendar year than it is now. Right. We're going to be coming up with different.
Rajvir Madan [00:17:48]:
Yeah.
Rajvir Madan [00:17:48]:
And I think you're already seeing sort of some of the pullback, or people are starting to think about some of the pullback.
Rajvir Madan [00:17:55]:
Right.
Rajvir Madan [00:17:56]:
I mean, I'll give you the classic example now from New York Times. I mean, there's a big lawsuit from the New York Times against chat, GPT or against OpenAI, let's say, because they're claiming that their data was used to train the large language models that OpenAI uses. Right. So I think a lot of this is still, we're still in the first innings of a lot of this, and a lot of this still needs to be sort of worked out. And I'm quite interested in sort of looking at how some of the legal and some of the regulatory frameworks around all of this actually continue to evolve.
Steve Swan [00:18:32]:
It's going to be interesting. I don't know how they're going to sort through this. And again, that's way above my pay grade, so let somebody else handle it. So we hit AI. Anything else you want to add about AI that you think we need to get?
Rajvir Madan [00:18:52]:
You know, I think the only sort of other thing I would sort of add is, and this is a little bit of, maybe it's a little bit of a cheeky comment, but I'll sort of make it anyway. I think a lot of companies out there are sort of selling their tools as generative AI tools, and I would just take that with sort of a grain of salt.
Rajvir Madan [00:19:10]:
Right.
Rajvir Madan [00:19:10]:
Not everything is generative AI, and you don't need generative AI for a lot of the use cases that need sort of focused AI or precise AI or whatever you want to call that. Right. I think a lot of companies are sort of riding this wave of generative AI because anytime they say, oh, our platform uses generative AI, I think you got a bunch of people listening into that hype, let's say. But as a CIO, I often think about, so what if it's generative AI or not generative AI? I think so. My sort of point there is don't fall for some of that marketing tactic behind companies sort of calling their tools being generative AI enabled or not. I mean, you might not even need generative AI. And whether some of those tools are really generative AI or not generative AI is a whole different sort of discussion. So I think sort of just take some of that marketing that comes from some of these vendors with sort of a grain of salt.
Steve Swan [00:20:17]:
So I'm sorry to do this here. Right. If you want to give me a real quick, I don't know, snapshot for somebody that's watching this, give them a quick definition, generative AI, so that they know what they're looking at.
Rajvir Madan [00:20:30]:
Yeah.
Rajvir Madan [00:20:34]:
I'll sort of simplify it, maybe. Please. This is sort of the simplistic sort of definition that I sort of think about. I think it comes from the keyword generative, right? So generative is, I sort of think of it as is this model sort of generating content for me? That's when I sort of think of the notion of sort of generative AI, like, is it generating intelligence for me? Is it generating sort of content for me? Is it sort of putting together, let's say, different pieces of information and then sort of giving me sort of the best answer? That's sort of my definition of sort of generative AI. Right? Whereas when I sort of think of other models that have not had maybe that generative AI component, what their sort of sweet spot is, is often around connecting different pieces of information together that you did not think could be connected. So if you take this example of using AI for drug discovery and drug development, there's very little sort of generative component there. It's not sort of generating any new content or that kind of stuff. What those kind of models are doing is, I mean, think of like neural network type models.
Rajvir Madan [00:22:02]:
These are models that are like connecting different pieces of information together and able to sort of come to you with an inference as to whether this molecule, this drug, is really going to work and on which patients is really going to work. Right. So that's how I sort of think of it in a simplistic way in my head.
Steve Swan [00:22:21]:
No, that's good. Yeah, I just wanted to, and again, I apologize for doing that. Just somebody who's watching to be like, well, hold on, what am I looking at? What am I thinking about? Okay, so we've hit AI. What other trends do you see from a technology perspective for our industry, either here or coming or that sort of thing? Any thoughts there?
Rajvir Madan [00:22:40]:
Yeah, I think the other sort of big trend I see, look, I think there's trends that are starting to get to a mature state with things like cloud, move to the cloud, let's say. Things like cybersecurity, I think will continue to be important, and the use of AI to enhance your threat intelligence will continue to be important as well. But the other sort of big trend that I see, which I think is going to really help organizations deliver incremental value through the use of technology, is around automation. So I see a lot of sort of trends around automation, whether it's robotic process automation or other types of automation. Let's say that's one of the really big trends that I see in our space. And I think if we look across any organization, there's a ton of, I would say, inefficient processes.
Rajvir Madan [00:23:36]:
Right?
Rajvir Madan [00:23:37]:
I mean, this is the case across a lot of organizations that I study that I have been involved with and that kind of stuff. Right. And I think the use of automation to really streamline some of those processes, to make those processes much more efficient is one of the other sort of trends that I see. And you're starting to really see the technology sort of get there, right? And I think when I say the technology sort of get there, I had previously sort of seen technology that helped you automate inefficient processes, but now there's technologies that are even helping you identify what those inefficient processes are and then help you sort of automate and optimize those processes.
Rajvir Madan [00:24:21]:
Right.
Rajvir Madan [00:24:21]:
So I think that's a huge trend that I continue to sort of see in this space or in the technology field outside of AI. Look, I think others will tell you about trends like AR VR, or trends like the metaverse, for instance, or quantum computing. And, look, I think some of those I'm a little bit on the fence about, especially with the whole sort of metaverse concept. I'm on the fence about that. I don't know if it's really going to take off or not. I think there was a need for much more sort of virtual interaction during COVID and as we're getting more and more out of COVID I think the need for sort of AR and VR in a daily setting, I think is starting to probably die out. I think there's still need for it. With certain things around training or education, it would be good to use AR, VR to train sort of a doctor maybe, or give them at least initial training on how to perform some kind of surgery, for instance.
Rajvir Madan [00:25:27]:
Right.
Rajvir Madan [00:25:27]:
So I think there's still a place for it, but I think it was a little bit too much hype, in my opinion, and I think it's likely going to sort of plateau in terms of its usefulness at some point.
Steve Swan [00:25:44]:
Yeah, got it. I saw that as something like you said, that was needed during that time. Right, during COVID and such. And I'm sure the commercial real estate guys are pretty happy that it's dying off, right. They need somebody to pay their mortgages anyway. So I wanted to also dive in just final at the end here, a little bit about diversity, right? Diversity in your group and diversity recruiting and diversity in the workforce. Tell me about your department's diversity and inclusion recruitment efforts. Tell me a little bit about that and where you are there and how you guys go about doing that.
Rajvir Madan [00:26:19]:
Yeah, so, look, I think there's a few things that I'm quite sort of proud of when it comes to Arcudis and the work we do around diversity and inclusion. I think we are perhaps one of the only companies out there that have actually done a pay equity analysis across our entire company. And we've come out of that pay equity analysis with extremely, extremely sort of positive results in how at parity our pay was across multiple groups within the organization. And that's something that I'm quite proud of. I think a lot of companies sort of talk about that, but ultimately sort of don't end up sort of putting their money where their mouth is, quite literally here. And I think that we are quite proud of the work that we've done there. I became a facilitator of this campaign that was started by Google. It's called I am remarkable.
Rajvir Madan [00:27:36]:
And I think it's really about getting females and underrepresented groups in the workplace to build what is called their sort of self promotion or their advocacy skills. And I have realized through my career that men are better at sort of talking about their accomplishments and their skills versus females and or underrepresented groups. I think there are some cultural nuances there as well.
Rajvir Madan [00:28:07]:
Right.
Rajvir Madan [00:28:07]:
I mean, I grew up in an indian household where my dad always said, sort of do your best, leave the rest. And I, quite early in my career, sort of realized that that wasn't going to work for me. It was just not going to sort of get me the success that I was hoping for because I wasn't able to sort of advocate for myself. I have been running these sort of iron remarkable workshops across the organization, did it at my previous organization, have been doing it at my current organization. And I think what that is helping us do is really sort of helping us develop the skill of sort of self advocacy and self promotion. And then there's many other things we do. I mean, there's a long list of, if you look at our board membership, for instance, very diverse board. If you look at our executive team, very diverse in terms of sort of the people of color representation that we have on our board.
Rajvir Madan [00:29:10]:
Sorry, on our executive team as well. And I think the last thing I would say is, and this is perhaps the most important thing I should have actually said this up front, is ultimately, we think very hard and we execute very hard on diversity and inclusion for our clinical trials because this is where ultimately, we're in the business of serving our patients. And through diverse recruitment, through thinking about various population sets that can benefit from our product, we do our best to incorporate those diverse population sets into our clinical trials, thus representing a diverse population of individuals that are actually testing and trialing our drugs. And I think that's the biggest sort of area where I would say that's one of the biggest areas that we're most proud of because making sure that we have a diverse patient base is really important to our mission and to our success.
Steve Swan [00:30:16]:
Well, and like you said, go all the way back to the beginning of the conversation. Advil may work on me, but it doesn't work on you. Right. We came from different parts of the world. If we didn't have that, maybe that has something. Again, I don't know. One of those works on me and one doesn't work on you, but it kind of talks to that point. Right.
Steve Swan [00:30:36]:
If we're not doing that, then we're just getting one population, one patient set. And that's not going to work.
Rajvir Madan [00:30:42]:
Exactly.
Steve Swan [00:30:43]:
No, it's not going to work at all. Well, that's great. That's awesome. I didn't realize that about your company. That's good stuff, like hearing that. Well, so I like asking one final question, but before I ask that question, I always like to ask the guest anything that we haven't covered that you want to hit on before I ask my one final question.
Rajvir Madan [00:31:02]:
I think we've covered quite a bit. I think the one thing that I sort of think about is sometimes what sort of keeps me up at night is the danger of technology. And as much of a technologist as I am, I think the misuse of technology can be quite alarming, let's say, to the human race in general.
Rajvir Madan [00:31:36]:
Right.
Rajvir Madan [00:31:36]:
I think we just have to sort of think about how fast is technology growing and who are some of the individuals that could potentially misuse this technology. And I have never been a big fan for process and regulation and policy and all of that. But when it comes to sort of AI, I worry. I worry about hallucination rates. I worry about misinformation. I worry about what is this technology going to do to underrepresented groups.
Rajvir Madan [00:32:22]:
Right.
Rajvir Madan [00:32:23]:
And I think that as much as I am a believer in AI, I am actively looking at how AI can help my organization, and we'll continue to champion sort of AI efforts. I think we have to go into this with our eyes wide open, and if we don't go into this with our eyes wide open, I think we're going to fall flat on our face, and I think we just have to be careful about that.
Steve Swan [00:32:54]:
Well, yeah. So we all talk about the great usefulness for the science or the marketing team or whatever. But you're talking about there's a dark side. There can be a dark side there, right? You got to watch out. And especially with you hit on it real quickly, the automation of some of the things having to do with cybersecurity boy, again, there's a dark side of that, too. So it's a lot. And this stuff's growing fast. And to stay in front of, it's going to be hard.
Rajvir Madan [00:33:25]:
Exactly.
Rajvir Madan [00:33:25]:
Real hard.
Steve Swan [00:33:27]:
No, I appreciate it. Thank you. So the last question I like to ask folks, and I don't get you ready for this live music. What's been your favorite band? Favorite live band you've ever seen?
Rajvir Madan [00:33:41]:
So the favorite live band I've ever seen is Dave Matthews Band. That was quick. Yeah. Sorry if people were expecting me to say Taylor Swift. I'm not a swiftie. I wasn't going to sort of talk about Taylor Swift here, although I have an interesting hypothesis that Taylor Swift is going to decide our next election. So that's my sort of hypothesis. But I've never seen her.
Rajvir Madan [00:34:08]:
I don't think I'm ready to see her. But yes, I've seen Dave Matthews band a couple of times, and I think that was one of the best concerts I've ever been to now.
Steve Swan [00:34:18]:
So you like the Jam bands? Because that's what he falls into, that category.
Rajvir Madan [00:34:21]:
Right.
Rajvir Madan [00:34:22]:
There you go.
Rajvir Madan [00:34:22]:
Yeah.
Rajvir Madan [00:34:23]:
Jam, soft rock, whatever you want to call.
Rajvir Madan [00:34:25]:
Right.
Steve Swan [00:34:26]:
All right. So, like, dead and company, almond brothers, Susan, Tedeshi trucks, those kinds of.
Rajvir Madan [00:34:33]:
Good. Good.
Steve Swan [00:34:33]:
All right, cool.
Rajvir Madan [00:34:34]:
Yeah. Well.
Steve Swan [00:34:35]:
Well, listen, Raj, I appreciate your time. Thank you very much. It was great having know, and maybe at some point we'll do sort of a panel kind of thing at some point.
Rajvir Madan [00:34:45]:
Sounds great.