What Real AI Transformation Looks Like Inside an Enterprise

Transcript generated automatically from the episode audio and lightly cleaned up (filler words removed, misheard names corrected where clear). Speaker labels were added afterward and may contain errors.

Rob May: Hi everyone, I'm Rob May. I'm the co-founder and CEO of Neurometric.

Anna Kirk: And Anna Kirk, I'm sales lead at Thread AI.

Rob May: And this is the AI in NYC show. Our guest today is the CEO of a company at data called Datamatics and is an AI thought leader who has really driven the company's transformation from lots of IT services into AI. So we're going to talk a little bit about that. It's really really interesting really difficult thing to do. Under his leadership, Datamatics has garnered a bunch of awards including Forbes Asia's best under a billion. They rank among the top 1% of 20,000 listed companies across Asia and they've named been named one of India's top 500 value creators by Dun and Bradstreet. So Rahul Kanodia, welcome to the podcast.

So I want to jump into this and get started. I am fascinated. We have so many guests on the show who talk about the difficulty in AI transformation and this is something that you've been you've been CEO for over 20 years. So you've seen the rise of this and having Datamatics recognize now as an AI leader like what are some of the what are some of the bumps you had making that transition and what are some of the lessons learned that you have maybe advice for for other companies going through it?

Rahul Kanodia: So actually our AI journey started about 14 years ago. And we started with a small R&D cell looking at AI. Before that it was we called it AI but it was really not AI.

And this cell was about a bunch of 20 PhDs in computer science who were sort of doing R&D with it. And they had some very interesting breakthroughs but the real transition happened about 2 and a half years ago. So we were using ChatGPT 2.0 told before Microsoft invested in OpenAI and then therefore we were fairly well entrenched into that and then once Microsoft invested they sort of closed the doors they relaunched it with 3.0 when the whole world went wild suddenly. Yeah. At that point in time we launched a program called an eagle program and an eagle program we take about 30 or 40 top performers in the company and we let them lose. So we set ridiculously high targets for them.

Saying you've got to improve productivity by 100% in 3 months. Wow. And the only rule they have is that they can break every rule of the company. Wow. So the federal rule that is bold. That is bold. And we said AI is what we're going to do. So here's all the licenses that you need to have. Here's all the technology. Go and do whatever you want to do. So we just let them wire. And they really went crazy actually. Yeah, we didn't realize the cost of that that [laughter] it did hit us. But the point is that they brought in so many ideas and thoughts and proof of concept pilots projects that many different functions of the company started adopting them and that really brought about a mindset change. So as much as it's a technology change, it's more of a mindset change and that's really what set the Datamatics page. That's when it started changing and then then we said AI is of course the future and therefore everybody needs to follow the path of AI. So these guys led the initiative but then of course it permeated the organization and many others are following.

Anna Kirk: I feel like we talk frequently on the show about the concept of incrementality with innovation and I really respect the ability to be able to you know build like start small build something incrementally see where it goes. You don't necessarily have to bias towards having a topdown strategy. You can bias towards incrementality. So you know from your perspective I'd be curious if there are any other areas in which you feel that there are common misconceptions about AI amongst the enterprise community today.

Rahul Kanodia: I think there are some misconceptions. So one one is that they think generally customers feel that AI is perfect and it's about to 100% 100% accurately and perfectly and it's far from that. It's still very nice. Is still evolving and it makes errors so it can't be perfect. What I tell our customers is that look if you've got a 20 30 40% productivity gain that's great.

Take it home. Yeah. Don't expect 100%. So that's one. The other issue that they struggle with is they try to challenge the intelligence of AI. So they give the most difficult project and say let's see if we can solve it. But hello even humans struggle with it. [laughter] Yeah. So, so my suggestion is take high volume, low complexity, simple stuff, automate that. When it's high volume, that's when you have maximum financial savings. Yeah. And not when you do oneoff exotic little problem solving, right? So, very often they get into a different mode of trying to solve complex problems. And that's the wrong strategy.

Rob May: Yeah. Even even humans, you know, people forget like humans aren't perfect and 100% on these things. And so why would you expect AI to be? So there's there's this misconception that AI is this magic wand and you wave it and it solves all your problems.

Rahul Kanodia: Yeah. Nowhere close to that.

Rob May: Yeah. People don't understand that the models are probabilistic and that's really important.

Rahul Kanodia: Actually the other issue is that they also try to adopt generative AI. Problem is generative AI as you said is probabilistic. You need in many situations deterministic solutions and answers. Yeah. Particularly in the world of finance it has to be precise. So you need to have a combination of classic AI which is really ML, NLP, regression, things like that and generative AI and them working together. It's not either or.

Anna Kirk: Yeah. Doesn't necessarily mean you need to throw away decades of best practices you've already built upon.

Rahul Kanodia: Yeah. And honestly, sometimes even legacy systems work very well. You don't need to throw AI at everything. Yeah. If it's working, don't don't change it.

Anna Kirk: When you interact with your customers, do you advise them on some of the cultural and psychological changes need to adopt AI as well or is it primarily just the technology?

Rahul Kanodia: Both is more technology because on the culture part we can advise them but we cannot influence them. Yes, it's their HR team that needs to plug in and bring about their mindset change the culture change. So we do give them pointers and advice but beyond that we can't influence it.

Anna Kirk: How would you advise maybe the CEO of a company to begin in terms of enabling that type of mindset change within a company?

Rahul Kanodia: I think first of all you've got to allow many people to use it. That's number one. The second is that people try to deploy AI solutions on the periphery. One small thing here and a small thing there and honestly that never delivers an ROI. You've got to bring it into the core. You've got to bring it in and you have to integrate it with all your other systems. AI cannot be a standalone thing by itself. It has to talk to all your systems. So that element of integrating it with your other systems and bringing it into certain core operations is critical. And most CEOs are very nervous about doing that. So they still try to test the waters by testing the periphery and when they do that they say we don't see the ROI and when they don't see the ROI then they say okay it's not working. That's the problem.

Anna Kirk: That that brings up an interesting point. I mean, how do you have to garn you have to generate an ROI, but you also need to find a very datadriven way to measure that over time. How do you know in reality when leaders talk about getting an ROI, how how do you advise that they measure that?

Rahul Kanodia: Well, it's that's a little bit of a gray area. The reason it's a gray area is that suppose you save 20% of effort but that does not translate into a 20% cost saving because you still have to pay the person for 8 hours a day. So you're still paying full salary even though the person has saved 20% of his time. Yeah. And therefore you don't see the bottom line benefit. But the point I make is that even if you're not seeing any benefit, you still have to deploy AI. You should because when everybody's using AI, your mindsets are changing, your culture starts changing, the way you functions are changing and that as a is an equally important shift apart from just technology. Yeah. So I think when but when you do large large projects and high volume that's when you can see a 20 30 40% productivity gain can give you an ROI. But if you have one person working on one financial analysis, he says 2 hours a day is not going to save your cost.

Anna Kirk: Yeah. I also imagine that there's quite a bit of variance when you look at kind of the differences in time horizon there too. If you're expecting to generate immediate ROI over the course of two to three months, it might not happen. But when you look at the J curve over time, I mean I have yet to see anyone who has disputed that AI is so powerful.

Rahul Kanodia: Absolutely.

Rob May: Yeah. We've had a lot of different guests on the show from different industries, you know, nonprofits, fintech, obviously a lot of SaaS and AI startups, robotics. Like where if you look across your customer set and the places you're having an impact, what what industries or what use cases or types of companies like like what's the are there any patterns to like who's really really benefiting from it the most or the fastest?

Rahul Kanodia: I think those industries that are not heavily regulated but are very high in volume. So retail for example, logistics, travel these are high volume consumerf facing regulated ones are like insurance and banks problem with those is that and they are benefiting significantly mind you and we've got some solutions on agentic underwriting or agentic claim processing or loan underwriting adjudication. So they are benefiting heavily but there the risk is that if you make an error the regulators will come after you very hard. Yeah. So therefore they still need human in the loop. Ultimately a human has to sign off saying yes or no. If you approve a loan or approve a claim or don't approve it there are legal repercussions of that. So they are benefiting but they are still not able to shed the human because of legal responsibilities. Those that are not so critical with human approval. For example, if you're buying something online or you're purchasing an airline ticket, those are not so regulated and they are also benefiting and those are easier to implement.

Anna Kirk: I mean there's the beauty of kind of the non-determinism which can be great in so many ways when you talk about generative AI, but then also in the more regulated industries, you need both the more deterministic logic, more of the process automation as well as the human in the loop piece. I'm curious when people talk about AI agents, it's so buzzwordy. They're typically talking about, you know, more of the non-deterministic element, but how might you differentiate that kind of more agentic process versus, you know, the more traditional process automation RPA or some you know,

Rahul Kanodia: in some situations, RPA works very well and you really don't need to change it. You don't need to make it agentic. Agentic is used really when you're chasing an objective and it has to be autonomous. So when the AI agent is working by itself, it's thinking is reasoning and it's chasing an objective. That's when you and then there's and information is asymmetric. It's unstructured. That's when you need agentic. RPA does not work with unstructured information, right?

And so if you for example we analyze balance sheets and annual reports for companies that do credit analysis and we process about 15 million pages of annual reports every year. Now an annual report if you read is not structured each report looks different from the other right and that's when you use agentic AI whereas an RPA will be robotic in its approach that's what it is supposed to be and it'll pick up specific information post specific information but it does not interpret the information right and agent AI will interpret the information it'll reason it'll re iterate and challenge its own thought process and therefore come back with the so if you say what's what's the risk profile of this company? Should we give a $10 million credit basis the risk profile it can recommend yes or no? And that's where you need a because it has an objective and it can work fairly independently.

Anna Kirk: So I guess to that end where do you envision you know which business functions might have more of that unstructured data and might benefit more from you know an autonomous agentic approach versus traditional RP?

Rahul Kanodia: Actually all business functions have a lot of unstructured data. And then people are have not been used to it. For example, in your company you have a travel policy, right? And then you could there could be a change in the policy. You could be allowed to fly business class or first class for example or stay in a certain category of hotels. Now typically this policy would be captured in a word document. In a traditional mode you would go back and change the rules in the system. Now you don't need to do that. That you just upload the word document and it's done. This the system takes care of it. Right? So every function has unstructured information. It's just that we've been too used to putting it in a structured table or structured database for it to work. But now you don't need to do that. So it's across the board and you know a lot of information is being generated by all of us social media particularly and today we generate about 175 zabytes of data every year. That's a hell of a lot of data. That's a trillion gigabytes as one gigabyte. So with that kind of volume explosion that's happening, it's impossible for humans or any other system to process it. You need AI to process because it can interpret that information.

Rob May: How do you guys deal with lots of different customers and you know different parts of the world and everything else. How do customers view what's going on in the model side of the world, right? And specifically what I mean like what is their attitude towards open source models? Is that changing? You know, have the have the Frontier Labs done a really good job of sort of like branding and convincing people they're the best? Do they look at benchmarks? Like what are you seeing from the customer base and how they evaluate the models that you might have them use and how maybe how that's changing?

Rahul Kanodia: So there are some customers who want to have the latest and the greatest. The problem is that these models are quite expensive right and when you deploy it into a commercial environment actually you can end up paying it through your nose. Open source is about roughly 3 months behind or four months behind which is not too far behind. It's already there right? When the commercial pinch comes in that's when they look at open source and say let's look at alternatives and open source does a fairly good job. The problem with open source is that when you get into industries like banking or finance or they are because they're regulated the open source is not as secure so security becomes a big issue. So if you're looking at an industry where security is not such a big issue and accuracy is not so very critical open source or some lower-end models work very well. For organizations that are very sensitive to security if you go to any large GE example or any of these large companies then they would prefer to go with these proprietary OpenAI or Anthropic or Gemini or whatever.

Anna Kirk: Yeah. And I also imagine that IP retention is a big part of that and you know C CEOs might be thinking differently about that.

Rahul Kanodia: They do the fortunately IP can be protected quite well if you if you put in enough sort of you put a shell around it and you put guards around it. IP is not such a big issue. There's a perception that you lose IP. The perception is still there but in reality you can protect your IP quite quite easily.

Anna Kirk: M so when you're thinking about what type of role the CEO needs to take in this more kind of generative AI age what do you think that looks like and how do you think you know yourself being a CEO and also interfacing with many CEOs in your role what how do you think the role of the CEO is changing

Rahul Kanodia: I think the biggest challenge for the CEO is really to reinvent his organization how do you compete because the rules of the games have changed so much that your rules of competing in the market is very different. And therefore, how do you reinvent yourself? How do you restrategize? How do you capture that great opportunity? So, on the one hand, AI potentially has a disruption of about a $30 trillion disruption in the world. So, $30 trillion right, but then it'll create an opportunity for another $30 trillion. I mean, you see Nvidia, you see Google, you see the valuations of these companies, right? Anthropic, it's crazy. So, on the other hand, you're creating the opportunity as well. So, how do you capture that opportunity and unshackle yourself from the past which is in many ways over?

Rob May: Yeah. I mean, Jensen Huang, I think, says it's going to make, you know, world GDP like 500 trillion dollars, right? Like he's it's kind of crazy. Let's talk a little bit about the sort of responsible AI adoption, you know, how there's a lot of discussion. And I, you know, I don't know what it's like worldwide, but I know both in the US and Europe at least, there's a lot of discussion about regulation for AI and everything else. How much does that hit the company level? And that are are companies at a phase where they're just sort of like we're just trying to experiment and figure this stuff out or are they thinking about responsible AI adoption and all that stuff sort of from the beginning of their implementations?

Rahul Kanodia: Most companies a they're still experimenting so they're not so much onto that but many of them has very seriously thinking about governance before because they're so concerned about AI they're concerned about losing their IP they're concerned about the accuracy levels so up front they are trying to put in guard rails and governance in a way that AI does not go wild or run emo and because of that approach it they're building in the ethical components already not by design but by default just because you're doing all these garbage to protect yourself you are bringing in the immoral and ethical components in that so I've not seen any company or any enterprise go wrong with the ethics of AI where I fear AI can go wrong is not really companies but governments and governments can abuse AI I and there's nobody to correct them. A company, you can sue them, you can shut them down, you can penalize them, you can find them, what are you going to do to the government? Nothing. So and there are enough governments who will abuse it for their own benefit.

Anna Kirk: You talk about instilling guard rails within an organization and having the human in the loop. We, you know, often internally at our company, we talk about the difference between human in the loop and human on the loop. So, actually using AI to springboard your individual employees. That said, there are a lot of folks who are rightfully concerned about whether or not AI will be taking their jobs. Do you think that AI will be replacing jobs or just accelerating them?

Rahul Kanodia: I think it will accelerate them. It will replace some jobs. It'll create many new jobs. And I'll give you two or three philosophical answers to it, if you will. If you if you go back about 150 years ago, people rode horse carriages. Then came the automobile. It's replaced all horse carriages. You don't see any horse car on the on the road anymore. But there are more car drivers on the road than you had horse car riders then. And another example is if you go back about 300 years ago, the world was largely an agricultural economy. Agriculture got automated, mechanized, people moved to industry, manufacturing, plants, things like that got automated, people move to services financial services, healthcare services, IT, consulting, what have you, right? That is now getting automated on the back of AI. People will move to something else.

Now, in this journey, the world population has gone from 1.2 billion to 8 billion. You don't see joblessness. So 1.2 2 to8 billion is a huge explosion in people but you find more people doing more work so I do think it'll create many more opportunities but of it'll replace the traditional format with the new format like the horse rider is now becoming a car driver skill sets are different but fundamentally still moving from point A to point B

Rob May: yeah there's a I don't know if you guys have seen it but there's a there's a substack called AI as normal technology and the thesis I think it's two Princeton professors who have written this idea that like hey we think AI is going to be like dramatically different and there are some different characteristics but like there's a lot of evidence that it's going to play out similar to previous technology waves in terms of adoption and labor force impact and everything else. So that's something we should see if we can get those guided on sometime.

Anna Kirk: So you can talk about the effect of AI in aggregate on the workforce. You can talk about you can also talk about the effect of it within different generations and different seniority levels. What advice do you have for new like entry level workers who you know perhaps can't like the entry- level job now looks very different than it did 5 years ago. So what is your advice in this type of era where the rate of change is so accelerated?

Rahul Kanodia: I think one is that they need to embrace AI with both arms completely and understand how it functions. So therefore if you have expertise in computer science and AI engineering that would be immensely helpful whether you use it or not doesn't matter but it's the training of the mind. The second is to really think creatively because the way you would use these tools to deliver your job or your business is very different and therefore you've got to be very creative in your approach and therefore in college if you can look at creativity and being agile and nimble and really out of the box thinking and bringing that along with the knowledge of AI would be critical and the demand today is huge. So if any new fresh graduate is coming in into the work work space there's a huge demand for anybody who understands AI. So I would really jump into it completely heads first in.

Rob May: So it's a really really rapidly changing industry and so as you're sitting here as CEO of Datamatics and you're thinking about like your strategy you have to place some bets for your company on how you differentiate from you know the Accenture of the world who are trying to advise you on AI and the IBMs of the world and the frontier model labs and whatever like what are the bets that you guys are placing how are you trying to differentiate your own company and sort of like why have you chosen some of those those avenues is

Rahul Kanodia: I think the difference is that we are not approaching it from a consulting perspective but making it really work when it works and it delivers an ROI or a benefit that's when the customer buys in and in our in our company we work a lot with business processes and therefore we have hands-on experience in that and when you're deploying AI into a business process the core agenda is to make it work it's not about delivering a technology and then say okay it's over to you so and that differentiates us. So when we go live with many projects, the customers actually see a proper ROI and that's what gives them the confidence.

Rob May: What does that to the extent that you can share what does that look like in practice? Like what are some examples of common transformation?

Rahul Kanodia: So for example there's one company that we worked with where we called it talk to your data. They had data of 10 years across 170 countries and they trade in yarn and fiber and things like that right so they've got so many different types of yarn so many different types of fiber and this is global information so today if you want to buy it it it can tell you whether you should buy from Bangladesh and route it via China bases the duty rates because they've got duty rates of every every day for last 10 for every country. Wow. Now this is a huge database but there's no system. It's the AI talking to the data directly. So and you talk to it in natural English language saying tell me you know should I buy from Japan? This is the yard yes or no. And it gives you that analysis.

There's another company where they had about $650 million of debtors outstanding and they were doing some transport related activity. We were able to process the bill of lading and things like that and they came down from a thousand people to 400 people and the cycle time went from 8 hours to 4 hours and net net they were able to reduce about 20% of their debt is outstanding. So $130 million of cash flow.

Rob May: That's a lot of money.

Rahul Kanodia: That is a lot of money. Yeah. For any size company. Yes.

Anna Kirk: So you split you split time between the US and India for work. Are there any differences in you know cultural differences, political, governmental differences in how different groups around the world are thinking about AI and its importance and anything like that that you've noticed or not between the US and India.

Rahul Kanodia: We work in Europe, we work in the Middle East, Australia as well. The differences I found not in the country and the culture but in terms of the organization mindset and we found three distinct types of organizations across the world. There's one set that I would call them the trailblazers. They say you know use us as guinea pig. We ready to try anything. Whatever the latest let's do it. If it's not done before we are the ones we're going to do it you know because we want to set the standards for others to follow. There's another set of companies who say if you've done it show it to me if you've not done it then I don't want to talk to you. Yeah. Now the problem is that AI nobody's done it before. Yeah. So they are the followers. And actually surprisingly there's a third set and there are some very large companies there. They say AI is a bad word. If you use AI I'm going to throw you out of the window. Don't talk. And that I find the most surprising because I don't know why anybody would be so scared to even talk about it.

Rob May: Yeah. I mean at a minimum you've got to be even if you're not ready to adopt it, you have to be experimenting, I think. Right.

Rahul Kanodia: Absolutely. The key is not about how AI delivers. The key is can you deliver better than your competitor. Yeah. And if your competitor is half a step ahead of you, he's going to win. Yeah. Whether you use AI, it doesn't matter. So really, how do you get ahead of your competition more than how do you fully adopt AI? And they are everybody's adopted. So they are going to move ahead of you. Yeah.

Rob May: Flash forward 5 years from now. AI and NYC is the world's number one podcast asked in that time you know where where do you see Datamatics as a company like where do you see yourselves differentiating over the next five years

Rahul Kanodia: so that's a tough one so our approach to the market is to go with AI platforms and solutions so we talked about agentic underwriting agentic claim processing loan underwriting we are doing finance transformation for companies we are automating contact centers using agentic AI. So that's where I see it going as a platform based service or human in the loop kind of combination. 5 years the reason I find it difficult is because the world of AI is moving so rapidly. It's hard to predict one year ahead because literally every week there's something new when somebody's ahead of the curve.

So 5 years is more than a lifetime in this current era that we are in. So it's hard to predict but I do see the way AI will function will be very very different and I'll give you a small example today you cannot imagine your life without your cell phone it's your YouTube emails your messaging your the phone itself is probably only one or two% of the cell phone it does everything else beyond just cell phone now 25 years ago my father was given a free cell phone it was launched in India for the first time he was given a free cell phone and they thought he's an influencer, right? So he'll use it and then people will will follow. And two months later he gave the phone back to the company. Oh wow. Now they said see I have no use for the self pay because all it did was make a phone call and today it does so much more than make phone calls.

Similarly I think AI is so basic right now so rudimentary in five or 10 years it'll be doing so much more we can't even imagine what it'll do and therefore we'll be wondering. So it's hard to really say where it'll be 5 years but 5 years ago it was very clear that you're going to have a wave. Today we are in the wave and you can see the impact it's already having. So I think it'll be hugely transformative but we have to therefore be very nimble and adapt to the transformation that's happening literally on a weekly basis.

Rob May: Yeah. I we talk a lot about we've had talked to some guests on the show about this is kind of like the ultimate industry for black swans coming right. Just like out of left field, new technologies, new applications that is it's really hard to predict and so kind of all you can do is set yourself up to like be nimble, watch things, have exposure and be ready. It's really really difficult to manage through.

Rahul Kanodia: Yeah, some industries are are ri you know education will get disrupted, legal disruption, research huge disruption, finance huge disruption, consulting heavy disruption. So they will be disrupted majorly but on the same time it'll create many more opportunities and that's what you got to focus on.

Anna Kirk: Yeah. And that rate of change can be I'm sure quite intimidating for for some leaders. What is your one piece of advice as we can close this segment of the show? For leaders who may still be reluctant to embrace AI.

Rahul Kanodia: I think the key is that if you are reluctant to embrace AI, you'll be out of business. Right. Yeah. So it's a matter of time and therefore the faster you move the better off you will be. Yeah. Even if you don't deploy it you've got to test the waters. You've got to understand it. You've got to experiment with it so that you're prepared to it's like a rocket ship. Once the rocket ship has taken off you cannot jump onto it. It's gone. You've got to jump onto it while it's still on the runway. And we are still on the runway. So this is the right time to jump on it.

Rob May: Yeah. There was a so we were talking before the show about backgrounds and stuff. And so when I lived in Louisville, Kentucky, in the early days of the internet, there was a there was a PR firm there and a media firm that refused to adopt the internet. And they they so there they actually put up a web page and all it said is we will adopt the internet when pigs fly. And they and people, you know, and in 2000 that wasn't crazy. It was like, yeah, maybe it's a fat. We'll see. I think they left it up till 2006 before they finally had to take it down. But by 2002 2003 people are like they're losing business because they're not taking seriously like the new you know the new technologies.

Rahul Kanodia: So they still in business out of curiosity.

Rob May: That's a good question. I don't I don't know. I' I'd have to go look it up. Haven't been there in a long time. But cool. Let's move into some some New York City stuff. So you know you've lived here multiple times. Spent a lot of time here. How have you seen the city of New York transform over the years from a, you know, the way that it embraces or thinks about like technology?

Rahul Kanodia: I think New York has not been the hub of technology in terms of generating technology, but New York has been a hub of adopting technology. So, I think it's done a wonderful job. A lot of the large corporates have headquarters here or a significant senior management sitting out of New York. So all the decision making takes place here. The financial industry is very very rife in New York media is very rife. These are industries that would adopt technology very easily and very fast. And decisions get made because that's where the that's where the money is. So New York has been an adopter in that sense. It has been leading the world of technology but not in terms of an R&D but in terms of the adoption and the funding behind it and all of that. So yeah. So if you if you are in New York, you get a good pulse of what's happening globally because you know and being a financial center of the globe. It's just not the US. You're really getting a pulse of the entire world sitting out in New York.

Rob May: Yeah. I mean that's part of what I love about it. And part of the stuff that we try to promote with the show, right, is the idea that this is kind of the hub of applied AI, right? And by being here, I mean Silicon Valley is awesome. Lots of great engineering talent, little bit of a monoculture. Whereas here you have all these interesting not just like media and finance that you mentioned but you know you've got you've got decent pharma industry you do have software you have fashion you know you have a bunch of stuff and it's all come together and so it gives you a really good insight into like where the world's going next and how people are applying this stuff which is kind of cool.

Rahul Kanodia: Absolutely. And the real use of AI is very diverse across all these industries. So here you get a good spectrum of the entire way it's applied in different industries. Yeah. So that's really very exciting actually.

Anna Kirk: And we were talking a bit earlier before the show about how you yourself have ties you know both in Mumbai and then also in New York City. What are the parallels that you see between the two having spent extensive time in both?

Rahul Kanodia: So both are very fast-paced. New York is faster of course than Mumbai but both are very fast-paced. Both have the hustle and bustle. New York of course is a little more systematic. Mumbai is a little more unruly. Because that's India. Both are financial hubs. Both cities don't sleep at night. So there's a lot of commonality. In fact, the first time I came to live in New York, it was in the early '90s, 1990 actually, January, and I met somebody from Gainesville, Florida. Okay. And then we were we were talking about she says, "How is it for you for a cultural adaptation from Mumbai to New York?" And I said, "Well, I feel at home. This is how Mumbai is. And for this guy from Gainesville, Florida, it was a huge culture shock. Yeah. Because, you know, it's just so different.

Anna Kirk: Yes, it's a good point. Probably, you know, pretty drastic coming from Louisville.

Rob May: It's definitely I mean, it's even draic. Well, I was in Boston beforehand and even Boston's a great city. All big city amenities, but again, the pace is slower and the it's the only city I've ever been in where like on a Monday night there's like five things you want to go do, right? And like and sometimes you just have to stay home and rest and so it's challenging in New York the pace

Rahul Kanodia: these are all cultural centers you've got your Broadway shows here in Mumbai is the same all your cultural programs are there is a financial center it's a shipping center so it's the international melting pot everybody comes to New York similarly everybody goes to Mumbai so it's very multicultural in many ways it's and because it's a global melting pot both of these cities it is very vibrant is full of life. Yeah.

Rob May: Yeah. So, we were talking before you were you were showing us the video. You had a son recently get married. And we were talking about Indian weddings. Do you think AI is going to have a have a positive or neutral or negative impact on sort of culture worldwide? Right. Do you think that there's a risk that these models trained on everything maybe like homogenize us a little bit or or do you think there's ways to do it where we will be able to keep the things that are special about different cultures, different parts of the world, different groups of people?

Rahul Kanodia: Many many companies and governments are trying to train the models on their specific data for their country and therefore it is getting a little more culturally sensitive as the training happens there. The basic models have been built in the US. So therefore this very US-centric. Yeah, fully understandable. But as it permeates different parts of the world, governments are realizing that they need to acclimatize or culturally, you know, modify the responses and you see that happening. China is doing that, India is doing that, some African countries are trying to do that. The only thing is they're not ready to spend that kind of billions of dollars behind it, but they're trying to do that. So I think the cultural elements are coming in and that is important because otherwise the world will just become too boring, too homogeneous.

Anna Kirk: We were talking a bit before the show about your love for Broadway and jazz. What it what impact do you think AI will have on music and the arts? How do you think it'll change how creative?

Rahul Kanodia: It will change drastically. People will create music using AI. People will create videos. For for my son's wedding we my my wife's mother's no more my grandmother's no more and we brought them back to life to bless the oh that's amazing beautiful absolutely so that's all done by AI otherwise you would go to a studio and shoot but then they don't exist anymore but create using AI we use their authentic voice and so you see and you can't tell the difference so it will have a huge impact for my son's graduation my wife made a song completely on AI congratulating. So it will impact all these industries hugely here.

Rob May: Yeah. We I mean relate so related to this when we started the show we tried to use AI to generate the music to start the show and we like just couldn't couldn't get what we want. So we ended up we ended up having a human do it. Actually the keyboardist for Vampire Weekend lives here in New York and wrote the opening music for the show because AI wasn't quite up to par yet.

Anna Kirk: So you know I also I think there's going to be a little bit of a reversion of authenticity. Like what really makes us authentic. I think we're going to have to think pretty deeply about that.

Rob May: Well, this gets back to your point earlier. You know, you were talking about how people need to be looking at like, you know, these these high volume repetitive things, whatever. Like I think that's how you structure the jobs of the future. I think that's the way to structure and use AI, which is like take the work nobody wants to do, take the work people don't like, take the data processing and all this kind of stuff and you assign that to AI and you leave the more interesting stuff to humans. You know, the judgment pieces, the culture pieces, the stringing stuff together feels like just watching the industry grow, we're still a long way from AI being able to do that effectively because it requires so much context requires a little bit of an idea of like what does it mean to be human and what do humans care about? And so it's going to be really interesting to see like I feel like society's on this tipping point where like we could go the wrong way or we could go the right way depending on how the leaders at all levels, right? Corporate government you know decide to adopt it and roll it out.

Rahul Kanodia: Sure. Well, I think I think using AI people will do a lot more than what they're doing today and we not we're not measuring that. I remember when I was in college in India in school we were not allowed to use calculators. So we had to manually mentally calculate everything and once you use calculators you are so much faster and therefore you do a lot more. Similarly with AI, it'll do a lot for you and therefore you've got spare time to do much more and I think that's the key thing. I give you another example. If you use people to shovel let's say sand, right? You got 10 people shoveling sand and now you've got this huge truck that comes can that can shovel the sand in one scoop and you don't need people. But as a result, you can construct high-rise buildings at 100 stories which with which shovels you could never. Yeah. So you're just you're going to do a lot more just because you have the power of AI in your hand rather than fearing job loss.

Rob May: Can I hire Datamatics to build a virtual Anna so that we can [laughter] shoot more shows faster? And I don't think anyone knows what's going going on inside this brain.

Anna Kirk: No, sometimes [laughter] I don't.

Rob May: So we like to wrap up on a sort of New York specific question. You know, you mentioned some things that you like. Is there is there a place in New York whether it's a park or a restaurant or or anything like like that when you come back you really try to get to like you're like this is just one of my New York like go-to places that you want to plug

Rahul Kanodia: so I love to walk in the central park that's it's always very relaxing it's nice so that's something I really love I love bakery so I make sure that I always go to Zabar’s and get some babka over there so that's something I do and one thing I love to do is go to Penn Station because it's just so busy.

Rob May: Yeah, it is. It's exciting.

Rahul Kanodia: It's exciting. You just see the hustle and bustle in there. That That's New York, you know.

Anna Kirk: Yeah. Yeah. And if you can get a chance to pause in an environment like that, everyone is busy getting to something and if you're just watching

Rahul Kanodia: Absolutely. Yeah. Do you Do you have a place you go when you've been gone for a while traveling? Like you come back and you're like, I got to

Anna Kirk: Well, I up up up until I'm moving now, but up until recently, I lived right I lived on the upper west side, so on the other side of Central Park. And so the Jacqueline Kennedy Onassis Reservoir was maybe my that was my thinking spawn. I just go walk around the reservoir and anytime I was in a mood to feel off pensive. What about you, Rob?

Rob May: I mean I go to Flatiron Room and get a bourbon when you when I've been gone for a while. It's just like Yeah. It's like my hangout and it's so Yeah. So I love it.

Anna Kirk: Or when you haven't been gone for a while

Rob May: and it's you know I do. I go there too much. Yeah. [laughter] sometimes I go there before the show and just hang a little bit. But and then last question I'll ask you actually because we I meant to get to this earlier and we missed out on it but what advice do you have? So you sit on some corporate boards, academic boards, you know, stuff like that. Are they leaning on you for AI stuff? Are you know do you do you give them advice or or people like that not really paying attention to it yet?

Rahul Kanodia: They do. They do. The academic world certainly does. Yeah. As some corporates would like to understand more about AI and then they do lean on me or or ask say can you bring in some expert who can advise us and so they do actually they're more interested in knowing where it's going and what direction it's taking rather than implementing it but they do look for experts to come in and sometimes I do facilitate people from our organization come in and talk to them

Rob May: yeah cool well Rahul Kanodia CEO of data thanks for being on the podcast today.

Rahul Kanodia: Thank you. Thank you, Anna.

Rob May: Everybody out there, thank you for watching. Like the show, leave comments, rate us. Those of you leaving witty comments on TikTok, please keep it up. We're starting to pay attention to those. You're you're going to see a show soon. We're going to respond to you guys. So, yeah, thanks for watching.

Rahul Kanodia: Great. Thank you. The next