Full Transcript

Dhruv Sharma: Hi there listeners, today is 13th of February. This is our pre-Valentine's Day stream and we've got, we've got three very interesting guests for you. Like right away. We've got Sanket or Sanky, who's the founder of InVideo and then the, the gentleman behind the InVideo AI Film Festival. We've had them before. In fact, the returning guests. Great to see all of you. Thank you.

Utsav Somani: Awesome. So why don't we start off by recapping how was the Bombay Film Festival and what's bringing you to Delhi now?

Hardeep Gambhir - AI Film Festival: The Bombay Film Festival was really good. I think we never expected to blow up like it did. We were definitely expecting it to blow up decently, but ended up going really well. Bunch of people that came like left me, left feeling like the AI films are actually getting like pretty decent. We got some really cool partnerships out of it. We are now working with Euros now, obviously Sanket here. We're in touch with folks at Morphic, like who was our past sponsor as well. I think the craziest stat from the event that I can possibly give you, you can maybe dive in later, is like 80% of the people that participated in the festival ended up either like leaving their jobs to pursue other jobs that they got from the film festival, or they got funding to work on their own films from like a couple of studios, a couple of venture firms in the US. So I think to sum it up, that's probably the heaviest stat we can give you.

Dhruv Sharma: No wonder they say AI is dangerous for job security. One film festival said you're out of a job, but you have a new one.

Chandan Perla - AI Film Festival: And on an average, they're making close to like 40 lakhs per person. Someone who used to make like five to seven lakhs before a year, is someone right now making 40 lakhs for working just in what they like, what they experimented and then happened to become their full time thing now.

Utsav Somani: How does one choose this career? Like I mean, apart from the creativity, the technology skills and everything that's required, like I mean, tools like in video, of course, are making it easier. But suppose Dhruv wants to become a film festival person, like he wants to make his own movie this weekend. Like, what should he do?

Sanket Shah - CEO, invideo: Yeah, mostly, I'm sorry, I'm going to take that up. I would ask Dhruv, like if he had any inclination towards filmmaking, and if he did not, then most likely he cannot become a filmmaker today. And I think that's the more important part, because the tools really, really are centered towards people who understand the craft. To give an example, how do you think you want your opening shot to be? How are you going to establish a character? What angle can bring in what kind of emotion or what are you trying to do with that particular thing? Or where should your chin be exactly to give out what particular emotion? I think that control is what, along with the story, brings in a great, great film video ad. And those details are actually, they remain with the filmmaker, they remain with people who grew up thinking about films and cameras and writing. And I don't think that is going to change. What is really changing is those people are so restricted, they don't make enough money. If you are like a great director, you're probably going to make three films in a decade and people will call you lucky. And the kind of money you get paid to do that is actually way less. And those are the barriers that get broken. Because now you don't have to convince an exec at a large production house to do a film, you can do a film if you really believe in the story, and you have a little bit of guts. So I think those are the parameters that's changing. And the parameters around Dhruv or Utsav or me wanting to make a film is not changing. I don't think I will be able to create a film. And that's, that's not the goal, or that's really the path that we are moving on as a company for sure.

Dhruv Sharma: In other words, AI can't be a substitute for creativity or intent. I mean, it can empower you with tools if you have both of those things, but it can't really act as a substitute for them.

Sanket Shah - CEO, invideo: Yes. And creativity is, and I by no means, you know, I mean that, hey, you are not creative or I'm not creative. I think we are all creative people in our own disciplines. This discipline here is filmmaking, it is ad filmmaking. And I don't think we have that level of creativity there. Because if we had it, we would already be doing something, we would have a portfolio and we'd be doing ads or we'd be writing things that resonated with people. And if that's not happening, really, then you'll be able to create something which you couldn't, but that's not going to be primetime, that's going to be for yourself, which is a use case, but that's never going to be primetime. Mostly.

Utsav Somani: Like an act. I mean, AI is like an axe, right? I mean, more you use it, sharper it gets, but eventually it depends on how you use it as well. But coming back to the film festival, the Delhi one, it's happening next week, during the India AI week as well. And you're doing, I mean, a pretty fancy venue, Qutub Minar. So tell us the story behind this. What can we expect at the Delhi edition?

Chandan Perla - AI Film Festival: So the story behind venue is a lore. We initially wanted to do it at, when you think of Delhi, you think of Gateway of India. And we really wanted to do that when me, Sahadeep and Sanket were on call, we were like instantly were like, this is where we wanted to do. But permissions wise, of course, they didn't give us. But you know what? The second in the list that we wanted is Red Fort. We all grew up watching Red Fort, right? That iconic building.

Utsav Somani: Imagine organizing. First edition in Delhi, you're literally going for the stars, basically. Why not? What's wrong with trying, right?

Sanket Shah - CEO, invideo: Anything, anything we try to do, we have to go for the stars. Otherwise the world's going to kill us.

Chandan Perla - AI Film Festival: Amazing. Love it. Love that philosophy. We tried getting Red Fort and we actually got it. And then after visiting the place, we realized that it's too massive, that you need an entire city to run it. Like it's a massive, massive property. We were like, okay, definitely in the bucket list for any future film festival or an event. So we were like something which is like small, controllable and well, like a jazz band, of course. So that's where we were like, Qutub Minar is definitely the part of the list. And yeah, thanks to every single person like Abhishek sir from the Meti and also the Culture Ministry and ASI who have supported us. Yeah, incredible people. I think big, big support.

Sanket Shah - CEO, invideo: Big support from the government. I'm just like super thankful. Big support from the government, government messages at 1am that, hey, what's happening with the venue? Do you need any help? And then we'd call them back at 11pm and all calls are answered. And yeah, just I think it's important for India to wake up and everyone who is trying to wake up. I think the government is supporting today.

Utsav Somani: Imagine, it takes six months. India AI Impact Week next week. I mean, pretty much every big gun in the world of AI is coming, except I think Elon Musk. I think everyone is there, all the large models, everyone is there. So I think kudos to the government for pulling this off. I can't even imagine stuff that might have happened behind the scenes to make this happen. For sure.

Hardeep Gambhir - AI Film Festival: Yeah, 100%. I mean, this is the biggest AI impact, AI summit in the entire world. Really looking at the history of it, not even US has organized something like this yet. So it's good to see India actually being at the forefront of this.

Dhruv Sharma: Yeah. I mean, guys, the pace at which, I mean, AI is moving at light speed right now, which is also maybe saying, I mean, anyway, it's moving incredibly fast and the weeks that have passed between the Mumbai and the Delhi edition of your film festival, what's changed? Like what can viewers and people who are coming in expect to see different wasn't even a possibility in Mumbai?

Sanket Shah - CEO, invideo: Decades have passed since. So we at NVIDIA constantly keep experimenting with what's possible now, and we have been working with several production houses in Bombay, as well as LA, and just in October, I don't think I would have thought that AI video is ready for primetime, and let's make a feature film. All the tests that we are doing with all these production houses, they're all done in India. They are all getting passed. They are all getting that, why do animation? Let's do this. Let's create an AI film entirely. Let's do this hybrid. Let's not do a car crash scene anymore. And then there's so many people involved in filmmaking, like there is an EP, but there's also a DOP, who is director of photography, who would decide that, hey, will this fit in the film? Is this okay for the vibe of the film? Will people still get entertained? And all those judgment calls are actually getting passed. Filmmaking is a long process, but in the second half of the year, I think multiple, multiple films will either be hybrid, and you will at least see four or five films that are just fully AI, that will not lose entertainment value. You will want to go out and watch it, not because it's AI, but because you got entertained. And I think that's the goal, like just the value of doing an AI film, because it's AI, it's temporary, it's one time, it's two times, but getting commercial success and critical success is extremely important. And I think we might be two months too early, but that's it, we are almost there. It's actually a great time to start a film, because then you can swap a few things out and it will be ready by the time the world changes again.

Utsav Somani: Will Smith can finally eat his spaghetti.

Chandan Perla - AI Film Festival: I'm having a fight with spaghetti, which came like three days ago, and it's doing incredibly well.

Utsav Somani: But tell us a little bit about the copyright and the IP issues in this, like, is there, I mean, filmmakers who might skirt that a little bit to make a better product?

Sanket Shah - CEO, invideo: There are, I mean, IP and copyright are two interesting, important issues, of course. But if we look at it, right, like if your intent is not to really infringe something, you will not. Now, the model training layer in this two-layered cake, specifically on the model training layer part, I think that that's just something that we don't control. And I don't think we should think too much about it, there are, of course, two camps. The one camp is that, no, no, let's, what about the IP? And I don't know how they're training models. And the second camp is like, I do not want to really do an accident shot. I don't want to block the road. I don't want to crash a car. I don't want a car to go down a hill. I just don't want to do it, because it's just not feasible economically to do it or even practically to do it. Right. So there are two camps. The second camp is where I am mostly. But if you don't want to infringe something, if you're not trying to copy something, if you're trying to build something from scratch, you should be able to do that, you are able to do that. So I don't think that part is an issue. The model layer part of how the data is trained is, of course, a particular type of opinion people have. But I don't think it matters today. And I don't think we should just think about it today as just like normal humans, because our goal is to leverage the best technology and really grow.

Dhruv Sharma: You know, Sankey, even as you're saying this, I'm thinking of like legendary animators like Walt Disney himself. He didn't have AI at his disposal, but his imagination was, you know, like, I mean, there's probably very few people who walked the earth who had the same imagination as him and he could invent characters and tell stories that still moved you, even though you're a human being on the screen.

Sanket Shah - CEO, invideo: Correct. Correct. And that's the story. That's the story that becomes even more important. I think the writer in this world becomes the most important, like every real project that we are doing, real project, like I'm not talking about an Instagram read, every real project, the rigor it takes to get a writer and think that you can work with the writer to create an AI feature film is actually extremely, extremely important. So I have come to the conclusion that they're not really, the jobs are really not going anywhere besides like, of course, concerns about, say, a cameraman, but otherwise, like broader part of the creative jobs are not going anywhere. In fact, I feel the only things that are going to change is everyone's going to have more work.

Utsav Somani: A couple of questions for the India Film Festival team before we let them go and continue with Sanket. So what is the I mean, I have a couple of open ended ones, different directions as well. What is the average age of a participant at this India AI Film Festival? What is the program that you're following for the Delhi one? I believe you moved away from the hackathon format, which you did in Bombay. And where can people buy tickets for your film festival? And what's the best way to get invited?

Hardeep Gambhir - AI Film Festival: Yeah, for sure. For the first part, honestly, we haven't been asking people their ages explicitly. But my guess is like looking at the last Mumbai Film Festival and a couple of people we know personally from this, from this coming India Film Festival, I think 24 is like a good age. These people are usually like, in the lines of what Sanket said, these people are actually really like filmmakers previously who are like now pivoting and using AFM more.

Chandan Perla - AI Film Festival: So that's for the first part. And also one common trend that I happen to see is that it's cinematographers who have been able to do an exceptional job at prompting because they could imagine how to get it done.

Sanket Shah - CEO, invideo: Cinematographer is most likely the zone of DOP, but even director's assistant. Director's assistant is a role who are, they prompt the set, they are still prompting the set. Like internally, that's what we call it, like this person prompts the set, they're still prompting the set.

Chandan Perla - AI Film Festival: And for the second question of how do we see the flow as previously, we've done it in a hackathon model. But now this time we were like, you know what, let's give it more time to these creators and then let them cook the way they want it to be. So this is going to be more of like a global showcase, like a showcase of the best films that they've made both in India and also internationally.

Sanket Shah - CEO, invideo: Actually I want to just like add on to that, what do we really want to gather out of this? In my view, what you want to get out of it is try to simulate the real world. The real world doesn't have to be a hackathon. It doesn't have to be 100% AI. It needs to be AI leveraged. It's about the story. It's not about did they use InVideo or not. So we don't restrict them on using InVideo. We don't restrict them on saying that you have to have a 100% AI film. So those things are entirely up to them because it's really about the story they are trying to make, say, tell the world. So we just want to emulate the reality of what's happening in the world. And it's all coming back from those insights of all of us meeting the world and touching ourselves with the reality that, hey, hybrid is great. These shots can happen like this. We can use AI as tools to enhance. We don't have to use the entire film to be like that. We should use mocap. AI doesn't get you expressions. You can do motion capture using AI. So why don't you act in your room and bring that smile or bring that anger to the film so that it has emotion? So we just want to really emulate the real world pipelines of what would get the ground running and mostly become more towards economic success than critical success. Either of them are fine. But I think economic success is going to be extremely important in making every technology go primetime.

Utsav Somani: Awesome. Guys, where can we buy tickets? And Dhruv and I are getting our tickets from you.

Hardeep Gambhir - AI Film Festival: Yeah, yeah, 100%. Awesome.

Utsav Somani: Where can our listeners find the tickets?

Hardeep Gambhir - AI Film Festival: So the website is indiafilmfestival.ai. Through that, you can click on register for tickets and it should redirect you to a loom of one. Yeah.

Dhruv Sharma: Do you guys have any AI film critics coming?

Hardeep Gambhir - AI Film Festival: Not yet, but they're more than welcome to come.

Utsav Somani: Do this .ai domain. It's by this country where I think they get 20% of their GDP from just domain registrations It's insane. Yeah, 20% approximately, maybe even more.

Chandan Perla - AI Film Festival: By the way, I want to tell something crazy. Sanketh definitely has an incredible announcement to give at the film festival. So we'd love to see you, all of you. Yeah. Awesome.

Utsav Somani: All the best, guys. Have a good sprint.

Sanket Shah - CEO, invideo: Thank you. Thank you.

Utsav Somani: All right, Sanketh, you've given us your philosophy about using tools like NVIDIA, but let's explain what NVIDIA is and who are the people using it.

Sanket Shah - CEO, invideo: Yeah. So NVIDIA AI, which we moved to, and there's something with taking risks and taking bull beds that I think we have internally been able to do. I have always believed that best products are built at the intersection of what's possible just now, married to a user need. We started NVIDIA AI three years back as an agentic video creation platform. And at that point of time, the tool and the AI was at a place where we could serve someone who cannot create a video to be able to create a particular type of video. And that had been the focus for the last two years. Mid last year, we started realizing that it's getting ready for primetime. And while serving all these users who can never create a video, we have ambitions to build pipelines and we have ambitions to build tools for real filmmakers. The philosophy of those tools will be very different than the philosophy that what we have followed, but we believe that that is where the fun is. That is where our curiosity is taking us. And those are the parts where you have to go to. In the next two or three weeks, we'll have a decently large release coming out, which is going to be a tool really, really for anyone who is making a particular type of film. And they have different types of components, which is how's the character? Let's iterate on the character for a day. We don't want people to build a character automatically and build time characters automatically. And that is where you will often see that the filmmaker will take inputs from a costume designer of what do I want Dhruv to exactly look like in what scenario? What are the worlds where Dhruv is going to come in and Utsav and Dhruv are going to interact? What does that world look like? And you have to have continuity from that world, from the story, and you will iterate on the story and the screenplay and the characters and the locations. And then you get very specific, minute control across all these parameters, which includes directing photography, which includes changing something in the world, which includes even smaller things. Now they have become smaller, which is like, how do we apply lip sync in a particular way? And how can I have a particular opening shot? And I don't like the emotion here. I want a real actor to act and bring in that emotion to the pipeline. So we are building all those pipelines and the new versions will attract an absolutely different type of audience. But if I just look at it today, 10 million videos get created every single month. We have decent scale of 50 million, 60 million plus users from across 190 countries. But I think there's something about being willing to take bold bets. And I think these are, I acknowledge it just generally that I think these are bets where I will either die or I will win. And those are the bets that like that's the kind of culture we want to follow.

Utsav Somani: You're like Elon Musk of video. There's a near-death pivot also that you went through when you were stagnant at 7 million ARR and now of course you've got 70 and maybe even more as well. So talk to us a little bit about the mindset that you had back in the day.

Sanket Shah - CEO, invideo: Look, I think, see, and I think that has to do with the, like there's so much context in a particular business. But what happens beyond a point is you start saturating a particular market you thought about. And there are market dynamics that you have to acknowledge, even if you don't like it. Those dynamics once hit needs a particular type of a change. And those changes really are very, very scary because you don't know till the rubber hits the road. And I think we are taking one more bet on those lines at this point of time. And I am really scared about them. And I'm really also excited about it. What happened at 7 million is like, you know, and it was, that was my first or second time. This was few years back. I've of course grown from there. What happened at that point of time is no matter what kind of features you build, no matter what kind of optimizations you bring in, no matter the kind of changes you make, you are not going to grow beyond 25, 30% year on year. That is just the nature of the market. And I think every entrepreneur should figure out a way to recognize that. Now, once you have recognized that that's the reality, you cannot take a small bet. You have to take a very radical bet, which is very different. It is mostly against certain philosophies and then rethink everything from scratch. When we do something like that, we actually stop looking at all the numbers. We stop looking at everything. We of course maintain the product. We make sure that the product succeeds. We add things, but we go and think what needs to happen. There's a particular level of rigor that goes into it. But towards the end, after all the rigor and everything that's applied, you have to take a bet. You have to have conviction in the bet and you will never know how that bet is going to play out. That has of course served very well for the business. We did extremely well. And then we started realizing that from here on, this is not our next 10x. And I will never be able to get to a 10x very easily without making a large radical bet. And the reason why such things happen is also because the world is moving at a particular pace and then you have to merge both of what's possible. How is it possible? And we think that filmmaking, ad filmmaking, ads are just getting ready. We might still be a few months early, but that is okay. So we have taken that bet and that bet actually in some form also brings in massive amount of risks to the current business because you're no longer in a protective mode. You are in fact saying that we'll just take that bet. And you take that risk and you see how it played out. But I truly believe in this part that several great entrepreneurs say that in business, when you are right and you hit a six, you don't get six runs, you get many more runs. And it's that. And I think I at least have three more bets inside me at NVIDIA today.

Utsav Somani: I mean, a guest who came on the show was good. I mean, and kind enough to differentiate between a pivot and an evolution of the business. I think we were talking about Misho with Tarjun Malhotra, right? Dhruv, you remember that? So sometimes it's just an evolution. It's not a complete pivot to a different thing as well. But yeah, thanks for sharing. Yeah.

Dhruv Sharma: I do have a related question, Sanket, which is when things are working perfectly fine, right? Maybe you're starting, as the founder, you're starting to see some things that other people are not. How do you, I want to understand your founder psychology at that point and how do you get others, how do you get them to buy in? And that includes your board, you know, investors, shareholders, employees, and other leaders within NVIDIA.

Sanket Shah - CEO, invideo: Yeah, I think like all of those are very tricky spots. And then I think there are different styles that different people operate on. My personal style is being very candid. Like I would say that I think we are figuring life out. And I would just put it out there that that's what's happening. Few people might be that, hey, this is what we are building. This is going to win the world. It brings in a particular type of motivation. So my style is very candid, straightforward. That's how, that's just my personality. And basically that is how we would operate. And the second thing is, I think if you're not going to be happy along the lines of your gut and instinct, then you're not going to have fun. And then anyway, like it's not going to work. So I think it's less about the board, the investors, et cetera, et cetera. They are very important, of course, but it's less about that. And it's just your willingness to say, what do you want to do? Where are you going to have fun? So it's tricky, different characteristics work for different people. For what works for me is just being, calling spade a spade. In fact, I call like spade a spade far too much than what is sometimes required and should move like the other end of it.

Utsav Somani: I'm sure it gets you into trouble or maybe keeps you out of it. But let's talk about the technology a little bit. The models are evolving. I remember reading about ByteDance, SeedDance, 2.0 coming out, Dhruv was mentioning Elon Musk and the XAI hands-on, where I think they were talking about the imagined feature of their Grok product. So these models are evolving so rapidly, like Sora as well. And then, I mean, every other day, I think you see something which is new. How does a business, I mean, you're probably using all of these as a buffet, right? I mean, you offer them to your users as a buffet, but how do you help the user decide what's good, what's not? I mean, even for you as a business, something might be more efficient, something might be cost-pocket-friendly as well. So what are the decision-making factors when the world is evolving so fast, like every three months there's a new model out there?

Sanket Shah - CEO, invideo: Yes, there are many factors. The most important factor and where we do magic is to figure out a way of how can we really make the most out of the models available. It's extremely difficult to... It's a harness in many forms where there are so many techniques to use LLMs and to use the video models. And those techniques have to be used very contextually, and they have to scale in a particular format that marries the filmmaking workflows. What InVideo is, is a full-fledged editor controlled by LLMs on which the models sit, which makes sure that you can mostly create most type of videos, provided we can get the AI to work a particular way. Now, all these models, image models, video models, and LLMs, they have a particular way to juice out the value from. And that is where we have to do the magic, that what is the AI cinematographer going to do? What is the AI director going to do? What is the AI costume designer going to do? And will it enable you to... Will it bring you closer to your vision or give you choices that helps you clarify the clarity that you did not have to get closer to your vision in some format? And then how do you increase the iteration speed on what format? So it's not one image. It's a long film. And how do you get closer to feeling how the film is going to land much faster into the process so that you can iterate there? So how can we... There is an animatic. How do you create an animatic faster? How do you create a pre-production meeting deck faster? And what does that deck lead to working all of them together? And then if you have an action sequence, what kind of model do you use? And sometimes even for one... All these films have every short two or three seconds, four seconds. Of course, there are one hours and stuff like that. But otherwise, it's a fast card. We'll use multiple models at multiple places to make sure that happens. And then all of them have to have same color, same vibe. You can't really prompt yourself out of that. You need to figure out a way to extract maximum value from AI. And that is where we do the magic.

Utsav Somani: Are you running an educational institute as well? Dude, I think you'll make for a good professor if you can teach AI filmmaking online to students across the world.

Sanket Shah - CEO, invideo: Yeah, there is a good chance that we'll take multiple courses across the Indian universities on AI filmmaking this year.

Dhruv Sharma: Oh, nice. Wow. Sanket, you're obviously called InVideo, but I'm curious. How do you feel about other modalities? Do you think there is... Okay, let me just end my question there. How do you feel about text speech and other modalities?

Sanket Shah - CEO, invideo: I don't know what you mean by how...

Dhruv Sharma: Is there a future where they become a part of what InVideo offers to its users?

Sanket Shah - CEO, invideo: I don't think so. Not LLMs for sure, not speech, not too much. I am a big fan of focus. Sometimes I want to go too far and say that, should I further verticalize focus? Because look, I will be very candid and honest as an entrepreneur. I do feel often that I'm not able to do one thing well. I don't want to definitely try and attempt three things at once. I think if I win one thing, there is future of doing other things. But I want to stay focused. And now I think about it is also when I was at a million dollars in revenue, I was thinking about 10. And I thought that if I get to 10, great. When I got to six, I was like, how do I get to 50? When I got there, I want to recalibrate and say that, how do I get to 300? I am, to be honest, not thinking beyond 300, 400 right now. But I'm sure that once I get to 300 or 250, I'll decide, do I want to go to a billion or not? And that is when that will happen. Right now, I really, really want to get there.

Utsav Somani: Also, maybe one final question for you before we let you go to the world of filmmaking. What is the path to 150 million look like? Or are you going to save it for next week?

Sanket Shah - CEO, invideo: No, it is. I told you already that we are taking a very, very large bet that can be very disruptive to business. That can be extremely disruptive. We are going to win or we are not going to win. We'll decide there. And hey, if the bet doesn't work, I told you that I have three more bets that I can take. So we'll continue taking the bets and willing to die on that hill. Or really win big. I think we'll win big.

Utsav Somani: Naini, thank you so much for putting India on the world of, I mean, on the map for the world of AI as well. And kudos to you. Hope to see you next week. Thank you so much. Thank you for coming on the show. Cheers.

Dhruv Sharma: Thank you so much.

Utsav Somani: All right. Let's move on to our next and final guest for the show. Pawan of Fashionza. Pawan, welcome to the show.

Pawan Gupta - CEO & Co-founder, Fashinza: Thanks Utsav. Hi. Hi Dhruv. Hey Pawan.

Utsav Somani: Thank you so much for giving us the time. I see a teddy bear at the back.

Pawan Gupta - CEO & Co-founder, Fashinza: Yeah, I ended up working from home today. There's no, I mean, I can't do this podcast in the office. There's just too much noise all the time.

Utsav Somani: Do you work on a factory floor or, I mean, different factories that you manage, or is it mostly, I mean, a more cabin friendly environment for you?

Pawan Gupta - CEO & Co-founder, Fashinza: I mean, no, it's, we don't work out of a factory floor. It's still a corporate office. There's still like too much hustle and too much operations going on all the time. You'll probably see carton boxes behind me if I were to do it from the office.

Utsav Somani: So for our listeners, let's introduce Fashionza to them.

Pawan Gupta - CEO & Co-founder, Fashinza: So think of Fashionza like an asset light manufacturer for the apparel industry. And how we differ from the existing industry is we specialize in manufacturing for the fast fashion industry. So let's say H&M is a customer, a pioneer in fast fashion. They need quick turnarounds. They can't, they have to continuously update the designs that based on the latest trends in the market. And unfortunately, the existing supply chain that has been built for the entire fashion manufacturing is not agile and fast enough. That's what we enable. How do we achieve turnaround times of 60 days, 90 days, even 30 days. And hence allow the retailers to go to their customers with the latest trends in fashion.

Utsav Somani: And you had quite a turnaround story. There was some headlines that you're returning capital to investors.

Pawan Gupta - CEO & Co-founder, Fashinza: We did return a bit of capital. I mean, not everything, but did return some capital. I think the business went through a massive change. It used to be credit heavy at one point of time. And then like we did the entire cleanup and we said that, okay, credit is not something that we can be great at. That's left best left to do for the banks and the NBFCs. So how do we build out a business which is less reliant on credit, which is not taking too much credit risk. And hence like that meant that we had to rethink the business from the ground. So we did that. That turned out, turned around pretty well. We work with like four of the top 10 retailers of the world. So yeah, I think it's been a good journey.

Dhruv Sharma: Pavan, do you think we could spend a minute or two, just getting a sense of what the value chain looks like in apparel?

Pawan Gupta - CEO & Co-founder, Fashinza: We can spend like as much time as we want, because that's very cool to my heart. I think what people like, what's very exciting, interesting for people to know about the supply chain is right now, most of the apparel industry, if I leave the fast fashion aside. So most of your brands, your Tommy Hilfiger, your Florence of the world. So today we're sitting in February. So they are now currently placing orders for manufacturing, or let's say goods that are going to go into stores in October or November, or even like later than that. So that's how long the supply chain today is.

Dhruv Sharma: Like an eight month lead time.

Utsav Somani: How much? Like for a fast fashion brand, like say Zara or any others that exist?

Pawan Gupta - CEO & Co-founder, Fashinza: So, and that's pretty interesting. Like right now, February, so Zara would let's say be placing order 16 to be delivered 60 days down the line, or maybe 90.

Utsav Somani: So very different mindset. And what Dhruv was, I think, asking about is the supply chain. Like, I mean, let's dive into say somebody is doing Tommy Hilfiger. Like what does the journey look like if the user platform fashions up?

Pawan Gupta - CEO & Co-founder, Fashinza: What we enable with Fashionsa is that we allow you to react to the market. So that's why, let's say a Tommy Hilfiger is not a great customer for us because they've got their unique brand strength. They are pretty much, I would say, trend setters. While for us, let's say somebody like, I won't name any metronome again, they're a good customer. But let's say any other brand which is fast fashion, their business is reacting to the trend. If they see, okay, much easier to react to something which is going to happen 60 days down the line. So they see, okay, Taylor Swift today performed at the stage wearing an absolutely orange colored dress. A bodycon suit, 100% orange colored. You don't really need to tell people that all the Taylor Swift fans are going to ask for an orange colored bodysuit. How do you get to the market with an orange colored bodysuit fast? That's the trend they are trying to chase. Because if you go to the market six months later, it's dead. Like nobody's asking for it again. So you have to go into the market. You come to Fashionsa, let's say. With Fashionsa, what we do is we figure out, okay, number one, which are the factories which have spare capacities available? That's number one. Second, we have our own design and sampling in-house. Like we run those units, one out of India, one out of Bangladesh. So we know where the material is going to come from. We design the material. We design it super quickly. We take use of available material in the market so that we're not going into the entire cycle of manufacturing material from scratch, which is another long cycle in itself. And then, so we take the available material. We design on top of it. We design based on absolutely trying to copy as much as possible what is the exact demand out in the market. Create 20 variations of it. And then there is something called like a cutting marker. So it's like a CAD drawing of what we are designing. And the brand looks up online on the platform and sees, okay, this is great. 20 designs, great. Five are great. Five, maybe not. 10, chuck it. So we send out all these CAD designs. We send out the material sources directly to the factories which have spare capacities available and hence can actually deliver within those timelines. And the entire cycle of that eight to 12 months gets cut short to less than 60 days.

Dhruv Sharma: Well, the factories you work with, are they capable of delivering products at various price points?

Pawan Gupta - CEO & Co-founder, Fashinza: Well, that's another good question. Yeah. And what we realized is that that's where a lot of these data science and data analytics comes in. I'm not going to say AI, but I could say that as well.

Utsav Somani: So you've not bought fashion.ai yet?

Pawan Gupta - CEO & Co-founder, Fashinza: I think there's something brewing there. I'll name it out. I mean, so actually we do use AI. So I mean, jokes aside, we do use AI to figure out, okay, what's trending out in the market, what's trending on Instagram based on the image analysis, based on the sentiment analysis, and then go back and already keep proactively designing before even customers even ask for something. So like a Zara comes and says, okay, do you have orange color dresses? Because I want to do 10 versions of orange colors. They're going to say, okay, here are 20 designs which are already trending out there. We have designed them, everything is ready. Just give the order and it's going to go into production.

Dhruv Sharma: Could you help us contextually, is like at a given point in time, how many different thousands of designs are you deconstructing and preparing to ramp up capacity if the orders come?

Utsav Somani: And also what's the average utilization of factories that are working on the fashion.ai network across the year?

Pawan Gupta - CEO & Co-founder, Fashinza: Good question. So I'll start with the question first. So what we do is like, I mean, we know for a fact that, these are the 40, 50 customers that we work with. And these 40, 50 customers are going to get, let's say they're going to produce 2000 designs with us every month. That's probably what we know. So we're always ready with five to 10,000 designs every month, new designs. And we keep most of the inspiration these days comes from what's trending on social media, what are celebrities wearing, what's coming out of Netflix, all the places like that. So we keep scanning that, our algorithms keep scanning that. And then we come out and say, okay, these are the 100, 200 designs which are trending because of all of these channels and REI throws that out. And then we go out and let's say use technology as well as a human layer to design 5,000 variations out of those 200 trends. Very simple, we don't try to go too much, too off script. Honestly, like it also comes down to how much we can rein in the designers so that they don't work outside of what the data is telling us. And then on top of it, like, hence the 5,000 designs. Designs also mean that we have to create samples ready. Without a sample, it's pretty much impossible to go into production. And that's why we brought everything in house. We do all of the like thousands of samples every month. And then the customers then effectively have a ready catalog of never launched designs. These are not like already out in the market, but then they can come in quickly, choose what samples they want to go out to the market with and allows them to cut short the entire designing, production, product development, sampling life cycle. So that's the, like that time becomes zero for them because we're doing it proactively. And then comes the second step of finding factories, which are let's say operating at a lower utilization. And what you've seen is that most of the market today operates at a 30 to 40% utilization. That too on an average basis, there are seasonal cycles for sure. In season, it would be 70 to 80% utilization. Out of season, it probably would be zero as well. And this is where we come in and we take, start working with those factories, which are low on utilization. And hence when they start working with the 30 to 40% average becomes 60 to 70%, you can never go beyond that.

Dhruv Sharma: I have a question to ask here on seasonality itself, Pavan. So for instance, for the winter season, how do manufacturers like ramp up capacity? Do they actually follow met forecasts in the months leading up to the winter and try and take a bold bet?

Pawan Gupta - CEO & Co-founder, Fashinza: It's honestly the worst season for a manufacturer. It's impossible to predict. And honestly, that's probably one of the reasons. And at the same time, like for fashion, there's very little you can do in terms of fashion in winter compared to spring, summer. That's the biggest season. And that's where we, that's a small part of our business. That's probably 10, 15%, like not more than that. And also for the factories, it's a tough season. It's a tough season. What ends up happening though is that, let's say this year, the winter was fairly well. So the factories would ramp up capacity in anticipation that next year would be slightly bigger. And then suddenly next year, winter is late or just very short. So then the next cycle, they would ramp down the capacity. So it always keeps on going. There's never, never an equilibrium.

Dhruv Sharma: Have you figured a way out yet to get demand forecasts like straight from the retailers or distributors from the resellers and not the manufacturers?

Pawan Gupta - CEO & Co-founder, Fashinza: It's tough. And honestly, our model is also based on the fact that you don't need to do too much demand forecasting. So we, because we work with a network of factories, so you can always come and scale up and scale down. And that's one of the reasons we have, we do not really have our own factories. We might like have always, like probably we're trying to get 20 to 25% of the capacity. And now that means like having our own units, but just the fact that we give the flexibility to the retailers so that they can react to the demand without committing anything is the model which becomes successful. That's one of the reasons we tend to shy away from taking those forecasts.

Utsav Somani: Nice. And let's talk about our neighbors for a little bit. So India, I mean, I don't know if I have my numbers mixed up, but India does 15 million of exports in a parallel Bangalore is probably 40, 50 and China is approximately three X of that. Politics changing in the world. I mean, crazy things happening on a daily basis, tariffs and of course wars, and like just problems between these nations as well. And amongst like the counter, I mean, buyers of like these nations products in different countries. So what is really happening? Is India gonna benefit from this? How do we become 15 to 100?

Pawan Gupta - CEO & Co-founder, Fashinza: Yeah, I think like we are betting heavy on India, I would say. And I think biggest reason for this is right now what we see is India is... So by the way, we're present in both India and Bangladesh. We don't manufacture out of these two geographies. What we see is that there is a strong trend in China to even for the government to move this industry out of China, the garmenting industry. And the biggest reason is it's actually pulling down the average GDP per capita for China. The wages in this industry are like not enough for China to move ahead when it comes to increasing the GDP per capita from let's say, I think it's about 12,000 today. And if they have to move it to 20,000, this industry will not help it. It will in fact be negative for it. So it's been, it has stopped a lot of stops for this industry. And it's trying to try out a lot of automation. But the challenge with this industry is because the value add is lower, very heavy labor driven thing. And then you have countries like Bangladesh and India, which has plenty of this low cost labor. So there's no incentive for factories in China to really automate. They would rather go and automate semiconductors or even like the robotics and cars. So that 100 or $200 billion of industry that is there in China, if you count the domestic as well, is going to move out sooner or later. How much of it? Like obviously we can all debate. It's not going to move out in the next 10 years, the entire thing. But even 20% of that is about $40 billion. And that $40 billion is equal to Bangladesh and three X of India. So India and apart from China and some of Bangladesh, India is the only geography in the world which has the population level at that price point, which can actually handle that scale. So hence, one of the reasons we are betting heavily on India, domestic industry is also massive for India. It's almost like $70, $80 billion for domestic itself. So the scale is always going to be there even if the exports suffer. So that's one reason we are betting heavy on India. And then second is just like the political stability is very high. We've had the same government for the past three terms. And even successive governments, transitional power in India is generally very peaceful. Nobody really likes to disrupt industries. And people are realizing that after the recent all the turmoil that has happened. I mean, even with the tariffs, Indian factories did not shut down. They were still running. Maybe the volume decreased a bit for the US, but then Europe kept them going. So it's like inherent stability in the country, which is what the large retailers especially are now really respecting. The challenge, obviously India has its own challenges in terms of not having large scale capacities, not having large scale labor mobilization in the single place. People are still scared of that. But I think some of those things are easing out. And we believe that with the right capacity and right quality factories in India, there's no dearth of demand.

Dhruv Sharma: Well, I'm also curious to ask you about textiles, right? Because that's upstream of apparel manufacturing. Do we have an edge with natural textiles? And for that matter, even synthetics. I'm given to understand that because we have, you know, large petrochemical refining operations, we also have, you know, a byproduct of that is, you know, is material that you use to make synthetic textiles. So can you, you're the best person to talk about those things.

Pawan Gupta - CEO & Co-founder, Fashinza: Yeah. So to be honest, like we lost the race there. And we are heavy into cotton. This is our mainstay. For some reason, we just stayed in cotton. For decades, I think nobody wanted to.

Dhruv Sharma: Which I personally say is a good thing, right? As they say, you live in natural fabrics and you perform in synthetics, but so cotton.

Pawan Gupta - CEO & Co-founder, Fashinza: You're right. And personally, I like to wear only cotton based products. And it's pretty good, cotton, wool, et cetera. But the economics are just so different. And not just economics, but for example, do you use, do you use Lululemon in your gym wear? You just can't identify.

Utsav Somani: Under Armour, Nike, I think the more traditional one. Lululemon, I think it's more yoga pants. So females typically use them, right?

Pawan Gupta - CEO & Co-founder, Fashinza: Don't tell me about it. It's massive in men's now.

Utsav Somani: I'd say it's massive in men's. I saw with his wife, I think runs a company called Bliss Club and they've done a campaign. About something similar to the Lululemon pants as well.

Pawan Gupta - CEO & Co-founder, Fashinza: Exactly. And like, it's got that stretch. It's got that like very extremely lightweight fabric, just like it's close to wearing nothing. And that's, that's the beauty of it. But you just can't.

Utsav Somani: Let's stay with that. Like, I mean, I've read a report that none of this stuff is actually healthy for you in the sense that it's not good to wear these synthetic fabrics on your body.

Pawan Gupta - CEO & Co-founder, Fashinza: I would say it's not, especially if your body is getting heat, which it would generate heat when in a gym or if you're running. So there's so many more studies coming in that polyester is actually harmful for multiple health reasons, not just your skin, but other. And there's chemical leakage that your skin absorbs from all of these fibers, which is effectively, it's plastic, right? I mean, it's petroleum. So it is true. And that's why I wear only cotton sweats when I even in gym as well, which sometimes I get stares at, but it is what it is. Like, I think that's more comfortable for me. And that's also like much more better health wise. But the challenge is that it's expensive. Like you can't do a lot of stuff with these guys who like anti-wicking, which is like, just like, which your, which your sweat out so that you stay dry, like all these dry fads. It's just impossible to make those kinds of products with natural fibers like cotton or linen. And hence, like you. So I would say for your regular gym wear, you don't really use it. But if you're a sportsman and you're playing cricket at a professional or a semi-professional level, then yes, I mean, there's a definitely definite difference in your performance. So for that, it's generally also called performance wear. That's how the entire category started. But yes, I mean, if your athletes are wearing those kind of wear, then yes, you also want to emulate that and wear those similar products. If that means you're wearing polyester, then yes, I mean, people do that. Although I would still say that the new age polyesters and nylons are, I would say, much better than the previous generations. But I would still agree with you in terms of natural fibers being natural. So, but coming back, like if the price is just absolutely, there's a huge difference. Like one of the things we do, so if we say, okay, we see a silk dress being very popular and on runway, and we see it's trending. One of our jobs is how do we take that silk dress, which is retailing at probably $500 and make similar looking, similar look and feel dress that can retail at $20. That's one of our jobs because fast fashion, that is the kind of customers we are looking at. So we go out in the market and we look for manmade fabrics, like all these polyesters and stuff, which can look and feel the same as the silk. And there's so many of them out there.

Utsav Somani: Take down the price basically and make it more affordable globally.

Pawan Gupta - CEO & Co-founder, Fashinza: A lot more, like it's a huge difference in price.

Utsav Somani: Pawan, we're running out of time. We've got to cover the news for our guests as well before we end the show. Any final 30 second question from you, Dhruv?

Dhruv Sharma: A 30 second, oh wow. I'm gonna clock now. So Pawan, maybe as a closing thought, just give us what the spectrum looks like, right? The spectrum of materials that are available to you, right? Like there's only one. I said 30 second with answer.

Utsav Somani: Dhruv has asked the whole spectrum.

Utsav Somani: No, no, Pawan, run us through that spectrum.

Dhruv Sharma: But I'll give you 30 seconds. So ultra fast fashion on one end and then super high end, whatever on the other.

Pawan Gupta - CEO & Co-founder, Fashinza: Yeah, so I'll start with it. There's luxury, which is like your Louis Vuitton, Gucci, Dior, et cetera, et cetera. Then there's your premium, which is more offloaded. Then you have mass premium, which is your Tommy Hilfiger, et cetera. And now Zara, for example, is pretty much in that category. And then you have the fast fashion, which is H&M. And then you have the discount stores, which in Indian context would be Vishal Mayawad. In US context, it's like T.J. Maxx or Walmart. And the trends flow from top to bottom.

Utsav Somani: And if you're wearing like Supima cotton tees, Dhruv and I love this Indian brand called March. Where would they fit into this whole thing?

Pawan Gupta - CEO & Co-founder, Fashinza: They would be somewhere around mass premium.

Utsav Somani: All right. Awesome. Pawan, thank you so much for coming on the show. All the best as you head into 2026.

Pawan Gupta - CEO & Co-founder, Fashinza: Thanks, Al. Thanks for having me.

Utsav Somani: Cheers. Cheers. All right. So just the eve of Valentine's Day, let's quickly cover the news just so that listeners can make their own plans for what the weekend holds for them. So Anthropic has closed a $30 billion series G at $380 billion post-money valuation. Did you read about this, Dhruv? I did. So this is just G? Wow. There's so many letters left. Yeah, series G. I mean, not many alphabets left. I want to see what actually happens. This is only second to opening as $40 billion round, which was larger than this. And what this means is, I mean, now there are so many investors which are overlapping between both the cap tables. I think it's a function of diversifying your capital sources as well. And probably none of these funds have the capability to take this AI war till the end or to the public markets as well. All this year, I think we see, we will test the retail public liquidity as well. But Microsoft, NVIDIA, Abu Dhabi, MGX, Altimedia Capital, where Sam was on the podcast asking to buy Brad's shares as well.

Dhruv Sharma: Sequoia. It reads more like the closing credits of a movie. That's how many names it has. So they've raised $30 billion at $350 pre.

Utsav Somani: And Alphawave is there as well.

Dhruv Sharma: Somewhat of an India connection to that. Indian origin, yes. Shout out to them. Great. But their revenue growth story is impressive. And then cloud code is dominating.

Utsav Somani: Yeah, I think so. It's become a hot favorite amongst the programming community where I think it's gone from literally 100 million basically last year in Jan to 14 billion right now. This growth is unheard of. They've gotten, I don't know what the multiple is, but I think maybe 15x.

Dhruv Sharma: If I'm not mistaken, I think it's either 14 or 20. We'll have to check. Even if it's 20, then it's at 19x. And so not bad.

Utsav Somani: Totally worth it. I mean, the output on cloud these days is just fascinating. And we talk about that article as well, the AI realism article, which I think you highlighted on our Slack chat. The wake up call, yes. The wake up call. I think that's very important. Let's talk about that actually for now. There are a bunch of things that we're skipping in the middle because Elon said to Anthropic on Twitter that your AI hates white cynicians. I don't know why he wants to get into every single fight, but I think I'm just happy. Anthropic is a better tool. And right now the favor is in their direction as well because I mean, the cloud code, the cloud co-work features, and just the whole tool. I mean, spend $20, get a subscription, and actually just skip one coffee and you will truly appreciate what you do.

Dhruv Sharma: Elon loves handing out tough love to competitors, friends, everyone alike.

Utsav Somani: I mean, a little bit of soreness there, but I mean, it's okay. I guess once you've achieved so much, then you have more fanboys, then fangirls, then people standing against you. But I think it's just, I mean, I think there can be some more taste.

Dhruv Sharma: But this article, and we'll maybe link it in our community and include it in our newsletter. It serves as a sobering reminder, a wake-up call almost. It's written by people who are, say, in the bubble for people, their friends and family who are not, that don't think this is gonna happen in the future. It's here now. No one is immune. Wake up, recalibrate your life. And yeah, I'd say it's not an alarmist article. It's actually very, very pragmatic. It's just a little ahead of its time. It's not an alarmist, yeah.

Utsav Somani: I actually read it after you sent us the link. And I think, I mean, they make really valid points that the world that we're entering into might be really, really uncertain, right? We've seen three years, I mean, Will Smith couldn't eat spaghetti from an AI video model. One of the modalities or the mechanisms to showcase the AI growth has been that Will Smith, as of yesterday, can eat spaghetti very cleanly and running twist in turn as well. So it's moving at a pace where all the IT stocks in India are down, just because India's primarily viewed as a back office country. So I think it's good for us that we use this time to upskill ourselves and use this technology to become better. Like, I mean, AI might not replace your jobs, but somebody who uses AI better than you might. So take all the time.

Dhruv Sharma: So even our message to people who are not in the tech community is invest in that subscription. Don't come close to the edge of possibility and use that to do more than just trivial things. Don't use AI as Google.

Utsav Somani: Yeah, don't use AI as Google. Like, honestly, like for now, I mean, even Nothing, that company that started by the founder of OnePlus, they basically have this tool now where you can actually make your app and publish it to your own phone. So you're basically coming to a world where you can literally make, I want to remind an app for taking my medicines at 12 o'clock. Can I create one? Like, literally, you can create an app and publish it on your particular phone. So you don't even need to know code. It's fascinating. I mean, it's natural language. You can talk via text tools, like 11 Labs and many others, like ChatGPD also now. It's fascinating what the world of AI holds. And India is doing a very big summit next week. Literally, who's who of this world. I don't think anything like this has been done across the world. Pretty excited. They're all coming down to India, except I think Elon Musk, there's Sam Altman, there's the Anthropic Dario, there's 11 Labs Matti, I mean, Alexander Wang from Meta. Like, I mean, it's a rockstar, like Jensen from- Exciting week ahead, yeah. It's exciting. Like, I mean, for India to pull off something like this means these people look at India seriously. We've been a consumer of all of their services. Meta gets around 600 million for WhatsApp from India. So I think it's going to be interesting to see how these companies treat India as a battleground. But there are signs that we might have companies- And also our homegrown leading lights like Sarvam, who they've been doing- And highlight what Sarvam is doing with the weekly drops, or the daily drops, actually, now.

Dhruv Sharma: Yeah, they plan like 10 days of daily drops. And I think the- Daily footing, I think, yeah. Sarvam Vision is the one that I think got a lot of attention, and rightfully so. That's their OCR model and capability, which is, it's trained in such a way that it does an incredible job of, in a sense, transcribing Indic language documents. And they think of it as not as a document extraction tool, but more as a knowledge extraction tool, because it's not just the words, but the context that's embedded therein, and sometimes even complicated things like tables and so on and so forth. So, you know, when OCR used to be really bad at serve, it would sometimes mistake I for one and S for five. You know, like- It's aging. Yes, I mean, I've used it for like seven, eight years ago, and now it's gotten to a point where it's scary good.

Utsav Somani: And all of this, I think, I mean, if you believe Elon Musk's vision, like robots will be at every home, basically folding and doing our dishes, all of that kind of stuff will also be automated away in the very near future. So making sense of all of these kind of data sources, I think it's exciting. But also what I think India can finally digitize all the scriptures that we've had. We've had, I mean, it's the land where Zero originated, Margot Dieter, and the things that have been highlighted by people on YouTube comments as well right now. So I think all of this finally comes to life. And it's exciting.

Dhruv Sharma: Like, I mean- We're chatting about this, right? Like a country like Russia, one language with 12 time zones, and we have one time zone, but we have 22 official languages. Maybe we speak 5,000, 6,000 dialects. So yeah, I mean, it is great to see what someone's come up with.

Utsav Somani: All right. Let's run through the next couple of things very quickly. So Bajaj Finance Scales AI. They've been using this company called Damana.ai where they've been making their loan officers pretty efficient. So they're transcribing all the data, which has been recorded. Whenever you call, get a call from these call centers with the cure, that thing that this call might be recorded for your training and audit purposes. So all of that finally is now being used to, used to like, I think makes sense. And they have 19,000, 25, 1925 crores in loans directly attributed to AI and 1600 crores from AI-assisted customer calls. So these are fascinating numbers. These are numbers that you mentioned on an earnings call and get spoken about. So this was highlighted.

Dhruv Sharma: But on a lighter note, they were hard to deal with even pre-AI, now that you're just going to be unhinged.

Utsav Somani: I mean, somebody actually said that once all of this becomes super cheap, like now you don't need a call center. You can literally throw up 10 agents in a cloud bot and they can make calls for you. The world is going to be insufferable when you start getting like 50 of these calls.

Dhruv Sharma: But you know what's going to save the, save the world itself? My agent is going to talk to your agent.

Utsav Somani: Talk to my agent. Remember that thing that people used to do, and they still do, EA is coordinating amongst each other on email. And I hated that. Like, I mean, it takes lesser time to actually just send a candle, leave the ego aside and just book that call. Good stuff happening in India, D2C as well. USP, which is a pharma company distributing several of my products and many other things in India, has acquired a 79% stake in Wellbeing Nutrition for almost 1600 crores in an all cash yield. Insane. Fireside owns approximately 20%. Hindustan Unilever owns approximately 20%. And the main founder of Niche owns approximately 51%. He's getting an 800 crore payout from this. And Wellbeing has demonstrated crazy growth as well. They had 120 crores and expecting 260 crores by FY26. So I think solid. They've gotten a good multiple as well. And their products look very, very nice also. Have you tried any of their products?

Dhruv Sharma: I'm yet to, but now maybe I will.

Utsav Somani: You should. I take a couple of things that they make. They're slightly on the higher side. Send me some recommendations, please. I'd love that. I will. I will after this one. So yeah, let's quickly end by fun things we saw on the internet. So Manish from TechX TechCrunch, he posted this thing on AI disruption medicines across earning calls transcript. And that's why the Indian IT stocks are taking a hit. Like they're getting disrupted by AI. So I mean, the funny thing is Narayan Murthy wanted people to work 70 hours a week. I think now people AI bots can work literally 24 seven. So I mean, something to note. And this is a wake up call. Be early, train yourself, upskill yourself, and the future might just be in your favor. Team size, talking about team size. So Elon Musk has made X very, very efficient. Nikita posted the head of product. Amita has 87,000 people. Microsoft has 221,000 people. Google has 190,000 people and X literally is run by, I mean, which is the town hall of the Internet is run by literally 30 people. It's just crazy.

Dhruv Sharma: You know, although that AI teams might also be really, really small.

Utsav Somani: Elon said that 30 people from X they also work on the feed as well. But I mean, the funny thing is like you and I were discussing before this are for you. Feeds are looking just the same. Like it's I mean, we're consuming the same information. So for us to surface exciting things, I don't know if people in our live streams are finding this exciting, but leave a comment. Give us a note if you think that we missed out something. And let's see if your algo is different from ours. Let's end by saying Google Gemini, everyday new model, ByteDance, SeepDance. Then Google had to, of course, do something else. Apparently only seven people in the world are better than, better coders than this model. Gemini 3 DeepThink for Google AI Ultra subscribers is out. It moved them to Fort Knox. Give them security. Dude, like I mean, yeah, seven people who can beat this. So yeah. All right. We've had a long show today. How do you feel about it?

Dhruv Sharma: I mean, I had a lot of fun. Like time passes by when you're having fun. But yeah.

Utsav Somani: Chalo. Have a wonderful weekend ahead. And thank you so much for tuning in folks.

Speaker 8: Bye bye. Thank you.