Top 3 AI Predictions For The Future

0:00Ladies and gentlemen, welcome to another video, and today we're going to do something different. I want to do some predictions for where I think the next 1 or 2 years will go. Now, the company that we run Airsoft is growing pretty fast. We're very well connected with the hiring ecosystem. We're hiring hundreds of people a month now, and I think we're learning a lot about what is in demand [music] and what is on its way out. And also a few things that we're noticing from the hundreds of enterprises we work with. And I thought what would be really useful for those of you watching is can

0:27we make some predictions that help you make decisions better? And here's a cheat code, right? I'm not actually making predictions. I'm just looking at where things are today that businesses are just starting to do right now, because those things will become mainstream in 2 or 3 years. And I'm making these predictions. So, hopefully it'll be useful for you because [music] the general populace lags a little bit behind some of these businesses making crazy decisions, and you can make career decisions accordingly. [music] Remember that you will disagree with a lot of these predictions. And I guess that's been true of the comment section in my channel [music] for many years, but over

0:55the years I feel like we've been proven right a lot. We've also made a few mistakes, but we're not like gambling on what the future would be. We're already seeing evidence, and then we're just pointing [music] out the evidence, right? Which a lot of people don't accept quickly. But before I get into my predictions for the next few years, I want to go into the last time I made predictions, and I had some predictions for 2025. And remember that none of what I'm talking about this existed back then, or they all existed in rudimentary forms. But let me read it out. My 2024 predictions for [music] 2025, and I want

1:24you to decide whether these came out to be true or false. So, here goes. My first prediction was that India will not just announce, but put out a playable demo of a AAA game. Well, you could argue on whether Unleash the Avatar is AAA or not, or AA, or whatever. Like, I think these are all marketing terms, but I think from a high-quality perspective, is this a

1:44high-quality game? And is there demos available? It's not just one. There are multiple. There are many people in India now trying, and many of them have, you know, at least [music] trailers out. We have a playable demo that we demoed both in India as well as with China. We're trying to get as much feedback as possible and even if I feel like we're still slightly below where we want to be, I know that we know how to get there, right? So, [music] at this time in 2025 when we put this out, zero. There was no one There was no There was You didn't even hear wind that

2:11anyone was trying to do something like this. [music] But, we had privileged information because remember we were trying to do this, we were talking to a bunch of people, we were thinking about should we even do it [music] and therefore we knew that not just us but a lot of other people are trying to do this and there will be some announcements soon. So, yeah, I think it turned out to be true. The second one is that content creators will absorb and be graded similar to consumer app, MAUs, DAUs, which is monthly active users and daily active users and will expand inventory [music] via AI avatars. This has turned out to

2:38be true. Content creators are now kind of seen the same as advertising businesses. Uh some of them gotten massive. We've seen some [music] of the really large comedians now go toe-to-toe with like media corporations in terms of revenue they're making and [music] some creators like me have expanded our inventory via AI avatars. At the time I put this out, we had seen a little bit of the AI avatar journey but it wasn't at this scale and nobody else had really done it, [music] right? Now, there are many creators across the world who have done AI avatars and many of them have expanded inventory. [music] They're making brand deal money on videos they

3:07wouldn't have been able to do otherwise because of AI avatar. The next thing is that VFX becomes trivial uh and it's just VFX is going to become iPhone app point and shoot. This has become true not just because of AI, which of course, you know, with C-DANCE [music] we're able to do it. In this time when I put this out, December 22, 2024, C-DANCE didn't exist. But, not just C-DANCE, right? Like, if you look at apps like Simulan or if you look at so many of the apps that are coming out now like Gemini Omni, right? I think, [music] you know,

3:33VFX is now pretty much point and shoot. The next thing I said was that video editing will go very mainstream. This has turned out to be true. We have like 500 video editors on payroll and uh we're constantly training thousands of video editors a year. The placement rates are through the roof. You know, the other day OpenAI had reached out to me and said, "Hey, you know, we really

3:48like working with your video editors." So, I think there are lots of video editing opportunities and they just continue to expand. I think it's gone pretty mainstream. The next thing I said was that Google and Nvidia will come out clear winners. And here, I think in 2025 end, Google looked like the clear winner. Nvidia was obviously was a winner [music] back then. It's a winner now as well. But Google, I think 2025 had a good year. 2026 looks like, you know, I think they're a little bit

4:10behind on the models. But as a company, as, you know, stock prices, I think they've done pretty well. Remember, people forget that YouTube is part of Google. But if you look at Alphabet or YouTube, I think they've done pretty well. If you look at Sundar Pichai's last analysis on their own quarter, they've done fantastically well. [music] So, even now, they're crushing it. And back then, nobody

4:27believed that Google would come back. Right? Back then, December 2024, they were pretty behind in the AI race. You know, Nvidia, sure, a lot of people could predict it. But back then, Google, nobody could. The next thing I said was that software founders will have an existential crisis, not directly because of AI, but because of competition and commoditization. I think software has gotten pretty commoditized, and I have

4:45more to speak about [music] it today. But yeah, I think a lot of software founders have pivoted away since 2024. And, you know, a lot of them used to disagree with me back then. Right? They used to be like, "Varun, I think you're wrong about this." We had debates. But I think most of them now, you know, definitely text me and say, "I think you were right about that era." Because see, here's the thing, right? We didn't make these decisions randomly. I was running a software company, and I was clipping

5:05the edge of this entire AI revolution. But it's not just AI. It was the fact that people were producing too much software. And every, you know, field in software had multiple competitors. And all of them were pricing below each other, right? So, the competition and commoditization of software really got very extreme. And there are really no moats, especially in an AI era. The next thing I spoke about is about 3D asset [music] generation. We said that 3D

5:28asset generation is going to get easier. And if you look at Mesh AI or Tripo now, they've gotten very, very, very good. When we started Unleash the Avatar, we couldn't actually use any of these tools because they weren't good. And to be [music] very honest, at this point now, our character work is already done. But if I were to do a new game, then maybe I'd definitely explore these tools and use them. Yeah, I just feel like it's very hard to change your game dev workflow in the middle [music] of things because these are large teams. It's very hard to pivot around too easy. But yeah,

5:52I think future game devs and everyone trying now will find these tools [music] that make the life very easy. You still need a few passes on it. You still need to do re-topology, but these tools have become very good. The next thing which I think I was wrong about was that always on board AI wire glasses see everything and help you during the day. I think we definitely have these glasses with the Meta Ray-Bans. They can see everything, but I think the limiting factor has been battery [music] life. I think you can't wear it throughout the day. You can wear it for a few hours when it's always on

6:18and I think we don't have any innovation in battery that allows it to run throughout the day. Unless you're willing to wear like a heavy battery pack around you, right? So I think while the technology definitely got good, I think the battery life didn't catch up. So I was definitely wrong about this one and I grade myself on these predictions every year. I've gotten better over the years and one feedback I give myself on is that any sort of prediction on hardware is not a one or two year prediction. I have to stretch my timelines out by five or 10 years. It just takes [music] time. Even if that

6:45technology exists, getting it diffused, getting it shipped all over the world, getting people to believe in it, the sales process, it takes a lot longer. Whereas with software stuff, everything takes a short period of time. Everything happens in like few weeks, right? Like things can change in months. But with hardware, it takes much longer than I thought it would. And the last one I said was like there will be lots of electric vehicles which I think we've got lots of new and pretty electric vehicles in the Indian market. Now, before we get to my predictions, I want to take a moment to talk about today's sponsor, Gamma, a platform that not only

7:14lets you create entire pitch decks from scratch, but does so with proper research, clean data-rich slides, [music] and a well-structured story, all within minutes. In fact, we even created a presentation for our video using Gamma and you can look at the results yourself. Now, this is a huge time saver for anyone who spends countless hours making presentations. Plus, you can even give it your own research, choose a template from their library, and let

7:41Gamma do its thing. [music] Anyways, considering 100 million users are already using Gamma, the next time you want to create a presentation and don't have much time, consider giving gamma.app a try. Now, let's go back to our video. Anyway, the top comment on this was "Varun, we're 19 months into AI takes our jobs. Devs are still here thriving and building actual tech, not just toggling background colors. How's the hype train running?" Well, now, and this is December 2024, when I'm in 2026,

8:10and I can tell you that we were right. And I think software has seen one of the biggest upheavals of time, and I just feel that what I felt at the end of when I was doing this product called Scenes in Avalon, I think everybody's feeling now, which is that anybody can build software. It's kind of gotten commoditized, and it's the same as the shift between writing machine code and writing some of these higher-level languages, which is that after a while

8:33you just forget about the machine code. Similarly, I think we've all kind of forgotten about code, right? And nobody's going in and debugging code. And this is not just true for average software engineers, it's also true for, you know, the best software engineers in the world. Even somebody like Linus, who didn't really believe in AI code in the very beginning, has changed his [music] mind on it. People like Notch, who built Minecraft, have changed their mind on it. And I think that brings me to my

8:54first prediction of the future. And honestly, this is the easiest prediction to make because we already see the evidence, and I'll give you the evidence as well. But I think software engineers, especially junior software engineering roles, are on the decline. And I think we're going to see like a big crash this year, right? The crash is already there. Honestly, I'm not really predicting the future, I'm predicting the present. [music] But some people haven't seen it, right? Like, especially if you're sitting down doing a job, some of this might seem sudden. "Oh, I suddenly got fired." or "I suddenly got let off." And I want to tell you the

9:22thought process and why this happens and the entire story because it's not just AI. It's not just this new technology came in and it's like competing with us because there's a lot more to software than just writing code. But I want to tell you the background process of this, right? And I want to tell you from my own experience [music] because this is the most critical thing that is important for you to understand. In my opinion, look, there's a single reason the software industry really boomed. And the idea came from Andreessen Horowitz saying from a long time ago, which is that software is eating the world,

9:47right? And the idea was simple, which is that if you're running a real estate company [music] and you build a house and you want to sell that house, the process of building the house takes time, the process of finding customers takes time, the process of selling it takes time, okay? And once you build a house, if you want to build a second house, the second house also takes just as much time. Maybe you'll have some efficiency, but it'll take 5%, 10%

10:08lesser. But software is different, okay? Software has something called a zero marginal cost of duplication, which means that if I make, you know, an app, if I want to sell another unit of the app to a different customer, it costs me nearly zero. So, everything beyond the first time I build the app, if it's the correct app and people really want this, [music] is kind of free. The cost is like zero. And therefore, uh you make lots of margin. So, while some of these companies in the real world, because of competition, go down to really low margins, there are lots of companies in India that run with like [music] 3%, 5%,

10:4110% margins, okay? Some companies that run in negatives or like very close to zero. But for a long time across the globe, there are software companies with [music] like 70, 80, 90% margin. And it's not just the margin that helped them become successful, the margin along with the fact that software is infinitely scalable. I don't have to do extra effort for the second customer [music] that comes in. Sometimes you'd write custom code, but like the truly scaling software companies didn't do that. They had one piece of software that they made everybody use. [music] And what they do is they'd go out, raise money and just throw money at getting

11:10more customers for their app. [music] And they lived on crazy amounts of free cash flow. So, what ended up happening was that engineers became really in demand, not because what they did was some mystical [music] craft. To be honest, I still look at somebody that solders things together or somebody that's able to build a road in the real world to be much better engineers than [music] software engineers, cuz the error margin, the risk that you are allowed exposure to when you build something like a Boeing 747, when you build a plane versus when you build software is so different. If you applied the same rules for engineering of a

11:39plane to software engineering, most of us would get fired cuz you have zero margin when it comes to building a plane. But, regardless of that, I think that the software engineers who of course focus on building apps made lot more money than these guys was because of the inherent scaling of [music] the business model. Two things happened because of this. The first thing is that many of these companies that really scaled, like the Google's and the Meta's and the Netflix, [music] they became what we call Big Tech, right? And they started realizing that hey, we have so much money, we are so profitable, and they all went IPO, and they said that

12:10instead of continuing to pay crazy amounts of tax, let's just hire more software engineers. Let's get optionality. Let's hire software engineers just to make sure our competitors don't get them, and let's hire software engineers and make them do cool projects. In fact, if you look at Google's entire research arm, or if you look at Meta's entire VR arm, in my opinion, the VR arm of Meta has barely made money. They've burned billions of dollars, but it's okay for them to burn billions of dollars because they've used money from their core platform, their

12:35core advertising platform to do that. And if you look at Google's entire purchase of DeepMind, it came because of their deep cash flow reserves. So, all of these companies started either buying other companies or they started like hiring good software engineers to keep running research experiments. Some of these projects really succeeded, like Gmail, or Microsoft's case, everything

12:52they've done with cloud services. [music] Some of these projects succeeded, and some of them, like Google Plus, failed. So, one of these two things happened, but it didn't matter because they continued to hire good software engineers from the market at good prices because they had the margins. Then, a second thing happened, which is VCs, venture capitalists came in the ecosystem, and they started saying, "I want to replicate some of the success of Google and Netflix and Meta. We should [music] invest in companies." So, all of this high growth, high salary stuff was happening with Big Tech, but a new wave of smaller companies without the revenues started raising lots of money

13:26on the promise that one day they'd make money. For example, this short video platform in the US that raised billions of dollars even before they had paying customers, right? And what would they do? They would say, "Hey, we have to compete with big tech. We have to compete with this other platform. Let's just throw money and get like the best software engineers. In fact, we have to compete with salaries from Facebook and Amazon and Google and whatnot, so let's offer them higher salaries." In fact, that's kind of how OpenAI grew as well, right? So, all companies in the venture capital space, they don't care about money in the early days because they're

13:53like, "We'll just take it from VCs." Which is a fancy way to say you're kind of borrowing it. Now, technically, you're not borrowing it because you're not liable to return it, but it is borrowing it because you're paying with equity. And somewhere in the middle of both of these phenomena, which is big tech paying software engineers to do cool interesting projects, uh sometimes even just sit around, and the fact that, you know, venture-funded startups were paying software engineers crazy money without [music] having real profits, I think this entire idea of software engineering broke. Because what started happening was my friend in this other company is making X,

14:22so therefore I should make X without looking at what the source of that wealth is. There was no judgment on why am I being paid this much. And I think what ended up happening was there was a delinking between what is the value a software engineer provides a project uh versus the amount they were compensated. You're paying between 25 lakhs and 50 lakhs an annum for one React engineer at that point because everybody wanted engineers because everybody was trying to hyper-scale their apps. But I think after that, definitely a correction happened where everybody in the ecosystem learned the hard way there isn't really space for so many apps. And

14:52software engineers, like any other job role in the world, their comp needs to be linked to how much revenue they're generating or helping generate. And I think somewhere in the middle software engineers forgot this. So, I think between 2021 and now there's been like the steep correction in this idea that, you know, we just pay people a lot because everyone else is paying them a lot. So, I think this difference has

15:11started to now really come into light. And that's where AI really breaks the equation because for a lot of these smaller startups that don't need the best software engineers, that can't afford the best software engineers, That are not trying to build the next Google, but something much simpler. I think what has happened is AI wipe coding has

15:28really replaced the software engineer. It's really reduced the cost of experimentation. Young founders are no longer raising as much money or even raising money because they realize raising money is kind of like debt. It feels cool to say I'm a funded entrepreneur, but it's kind of like a chef, you know, bragging about getting vegetables. You still have to cook, right? And I think now, you know, the coolness factor of raising money now looks like, oh, you're under debt. It's not really debt, but most top entrepreneurs in India kind of see it like that now. When I see a company that's raised like 20, 25 million dollars

15:56[music] and they have no revenue, in my head I think they're sitting under 25 million dollars of liquidation preference. If they ever sell the company, they'll have to first return this 25 million dollars back. If they don't have revenue or they don't have profit, how are they going to do that? So, a lot of these newer founders have become sophisticated. They don't want to raise VC as [music] easily. Venture capital has also got a

16:14little disappointed with software. Except for a few crazy outcomes many, many years ago, the outcomes are starting to windle now. They're starting to reduce. And also the compounding factor of, you know, having an Infosys and a TCS now start moving towards [music] wipe coding tools and the salaries there being kind of fixed. Now, there's another problem here, which is India looked at all the success and all these software engineers went out and made videos saying, "I'm making this much money and that much money." Which inspired every kid to be a software engineer. Five years ago, I had made a video called learn to code because I

16:42believed [music] it was the most important skill for you to learn to have a good career. In fact, software is the one true equalizer back then. If you learned how to write code, it didn't matter which college you were from. It didn't matter what you did, you would find an opportunity. But I think that era is coming to an [music] end and we need to recognize when these eras come to an end. I think intelligence is not about how [clears throat] well you can write code. What skill is valuable keeps shifting over time. [music] Intelligence is adaptability. Are you willing to change your mind in the

17:08presence of new evidence? If you have strong evidence and the wipe coding tools are constantly getting better, then you must be able to change your mind. And for me that moment happened when we built something called Auto Code Pro many years ago, where we realized that as the models keep getting better, they're going to be able to wipe code tools. And back then, it's going to be simple stuff like buttons and [music] simple web pages, but now it's whole apps. In fact, even till last year, even with wipe coding existing, I would still go to an engineer in the company and give them tasks. Today, I just do it

17:32myself. I just open Claude Code, costs me like 50 to 100 dollars, you know, expand my budget for the day, and just go out and build whatever I want, right? And there's so many tools we wipe code internally that are not trivial to build. It's no longer just CRUD apps, right? We're building complicated stuff. It's also helping me Codex and and Claude Code can now control my entire

17:49computer, so they also deploy it for me. I have a Railway account, they automatically deploy the apps for me, [music] and I really don't need anybody's help, right? So, I think the era has definitely changed. And if you look at my comments from 3 years ago, I would get absolutely thrashed, and people would say, "Bro, it can only do buttons." Last year, it's like, "Oh, it can only do web pages." 6 months ago, it was, "Oh, it can only do non-complex apps." Now, it's like, "Oh, it can only

18:12solve 10 unsolved math problems." Right? So, in every industry, and I never thought mathematicians would get upset that some of this is going away because I thought they do it for pure curiosity, but remember that all jobs have the element of the craft, right? You feel satisfied and you feel good because you are applying your craft to something, and if that gets taken away from you, you're of course going to be upset. So, it has happened to everybody, but remember that the world really rewards adaptability. It does not reward sticking to your beliefs when the evidence has changed around those beliefs. And you can attack me as much

18:42as you want for it. You're fully allowed to, you can say it in the comments, "You don't know anything, this, that, whatever." You're allowed to do it. But for 3 years, we've been saying it, and for 3 years, it has been truer [music] than ever. And it's almost like today, you can actually just download Codex or Claude Code and try it yourself. Most of you have, if you're an engineer, and you have to really ask yourself, if you say things like, "Oh, it still can't do

19:00this." or "It still can't do this." Astra, OpenAI Astra, which is the one that actually solved the math problems, or Claude's new model in training, or Gemini K3, which has now picked up on the open source front. I really do believe that this is not a wise industry for you to be in if your job is to purely write code. Now, if your job expands beyond writing code, if your job is to go out, build relationships, hire people, sell the product, find out what people want, be responsible for the entire package, be accountable for the outcome, then of course you're going to do well. And that goes back to the point

19:29I'm making which is that the entire AI wave now will affect you mostly if you're junior because as a junior you probably don't have these relationships. You probably don't understand what customers need as well. You probably aren't exposed directly [music] to sales. I told you there was an accountability gap between what software engineers did and how the company made money. Because of the entire VC thing, because of the crazy money in Silicon Valley, it got disconnected. People forgot. People start basing their salaries based on that other guys being paid that much, therefore I deserve to be paid this much. But in industries like this, when this happens, you have

20:00to go back to where is the value you drive. [music] So, if you know exactly how you drive the next 3 crores of revenue and you're directly responsible for it, then absolutely you'll make a lot of money. And a lot of senior software engineers actually know exactly how they contribute to revenue. And look, I have full empathy, right? I understand that you probably did a degree in something, you probably spent the last few years of your life in something. And there are

20:20two kinds of people in the world, right? People who hold on to their beliefs because that hope keeps them going through the day and people who understand instinctively that the world is tough. 200,000 years humans have fought for survival. If you watch any nature video, right? It'll be an animal running for its life and another predator like chasing that because it's hungry and it's going to die of starvation. That's nature, okay? So, easy is not a guaranteed. There is no guarantee and you know, I often see people and they'll be like, "Oh, this is bad government problem." But it's not the government's problem. Honestly, it's an adaptability problem. And I think if

20:52you don't realize that the world is tough [music] and adaptability is required of you, I don't think you're going to make it. I don't think sticking to your hope and your goal and your dreams, it's just going to get crushed. And for a lot of people like me with my startups, when I was younger and I had different ideas around what startups would work, when those dreams and hopes got crushed, I became a man. I went from a boy into a man after seeing a little bit of failure in life. Then realizing in order for me to make money, I have to provide value to the world. In order for

21:17me to make money, I need to know exactly how I drive value for an end customer. Right? Someone works with me but if they work with us on distribution, if we don't help them sell their products, it's not going to work. If we work with a company do YouTube for them, if that eventually doesn't lead to customers,

21:29that customer is going to leave us. Right? As an entrepreneur, you're not shielded from this, right? You know exactly what drives revenue and what doesn't. So, [music] you kind of learn it the hard way and I think it's a good thing because it makes you a mature, more capable part of the professional ecosystem. Knowing directly how you lead to revenue. And let me tell you one thing, right? It's 2026, career pivots are normal. I think we are against the idea of pivots because we believe that somehow we are doing something bad because we have to change what we were [music] doing, but changing your mind is the most important skill in

21:582026. It's always been the most important skill in fast-moving markets, but it's even more important than ever to be able to change your mind. I know this was a long piece about what I believe about the software industry, but I do believe we're going to see an absolute flattening in the next one or two years of the junior level roles in software, right? Especially if your job is only writing code. But for 99, no, 95% of the software engineers in India who are just getting by writing code and you are unclear about how your code leads to more money for the company, I'm sorry, but you have to [music] start

22:27thinking further than this. And please, for God's sake, don't hate me for saying this. Right? If we start doing la la la la, put our fingers in our ear and be like, "Oh, fairmonger. I don't want to listen to this." Right? When it actually arrives, you're not going to be equipped for it. By the way, it doesn't benefit me at all. Me introducing [music] this anxiety, brands don't want to work with me, you know, the audience gets upset. I don't win when I do this. There is no

22:50winning for me when [music] I do this. There's no benefit in me doing this at all, but I feel like it's my responsibility as someone who's been through [music] this, even if people on me to continue pushing forward and doing what I believe in and saying what I believe in. And I do believe that this is coming, knowing that it's going to piss off lots of AI companies watching my content. I don't care, right? I want you to do better because you followed me for a while, and I feel like it's my [music] responsibility to tell you this. I want to leave you with one stat. Like I told you, we're close

23:16to 600 employees today. 3 years ago, we hired 20 to 25 software engineers in that year. Last year, we hired maybe 10 to 15. [music] This includes game devs and stuff like that, right? So, C++ on that end. 10 to 15. In 2026, we have hired zero. [music] Zero. As a 600-man company that's known for introducing technology in content, that's known for doing crazy AI experiments, we've [music] hired no new talent. Look, we still have our existing talent. We've not fired our existing C++ game devs. We've not fired some of the

23:48guys who run Video Vault at EOS, right? But, we're not hiring anyone new. Especially coming from me, right? I've been someone who's been bullish about [music] software for a decade. To hire zero is a very strong statement. And look, whatever I say in public, [music] I think you should look at what I do. And if I'm hiring zero, it means this is where it's at. It's not like we don't build more software. We've actually built the most software we've ever built. Unreal Theatre, we have our own skill tree viewer that we've built ourselves. [music] We have our own CRM to do a lot of the stuff that we do at EOS, right? There's

24:16so much software that's wipe coded running inside the company. There's an entire animation [music] framework we have internally called Swiss knife. But, no software engineers applied on any of it. You take with that information what you want, but again, it's my job to tell you the truth. And I think the [music] first prediction is that I think in the next 1 or 2 years, we're going to see the collapse of

24:33junior-level software hiring. The second prediction is I think that there will be lots of jobs in India to train AI models. I want to give you a very good example. It's not going to be people like you and me generating this data. I think it is going to be sections of society we didn't really imagine to be contributing here. If you look at how the new home services companies are approaching this. If you look at how the Snapbits and the Prontos of the world are approaching this, they are putting a camera here and the camera's recording the maid do the work or the chef do the

25:05work. They're actually doing that, right? And they're storing this data. I don't know whether to sell it or to make the mod, you know, the robots themselves at some point in the future, but either way, I think all of us are going to be able to shoot this data with some of these glasses and then find a way to

25:19sell it or contribute it to a platform. See, the last time [music] these LLMs were made, we didn't know whether our content to be in the training data. My last book, [music] Pajama Profit, is certainly in the training data of all these platforms, right? And I wrote that book, right? This was pre-AI. We me and Abhinav wrote the book, but we didn't get compensated for it. But this time, I think everyone's aware that we're collecting this data for training purposes. So, I do believe that some of some people will participate willingly in this process. Lots of other people will shame them, but I'll tell you what,

25:46robotics is going to come out whether you like it or not. Maybe by preventing all these people from training it, uh you'll delay it by 1 or 2 years, right? But it's still going to happen. So, you might as well make money in the short term while it happens, and I do believe that uh we're going to see a lot more of this. A lot more businesses in India,

26:01factory workers are going to do this. They're just going to start recording uh a lot of hand movement as they get contracts. So, there'll be companies in India that give contracts to factory workers to just wear these glasses, right? They'll give contracts to, you know, every job, right? Like maybe I'm morning sweeper of the train station to wear something like this, right? This is valuable data. It helps these people make a little more money and there'll be a bunch of businesses that be the

26:22intermediary between these two. So, my last prediction is going to be a little bit of a niche one and it's not a 1-year long prediction. It's going to be like few years before this happens and it's a niche one and it's not going to be a popular one, but I'm going to say it anyway. Look, we've been spending a lot of time making a game and of course, you know, there will be improvements in game development. The tooling will improve, all of that. But I think the business model will change. It's fundamentally unstable for a game to spend $200, $300 making a game and then have the customer attention divided

26:51between hundreds of games that are coming out. In fact, per year, there are tens of thousands of games coming out from small to big. So, it's very hard to break through the noise. It's going to be extraordinarily competitive. And in some ways, it is moving exactly the same as movies. You know, I went on the Bad Decisions podcast and I was talking about this prediction that I have, which is that I think that games will move to entirely digital and streaming. And this

27:11was before Sony had removed the discs. And remember, we're a PlayStation partner, but I don't know this internal information that they're going to remove discs. But, it seemed like the obvious path. And there's a lot of people on the internet who are like, you know, "Discs are important, this and that." Which is true, right? Like, I would love for there to be a collector's edition disc of Unleash the Archers. We're racing to make that happen, actually. We're racing to see if we can be one of the few third-person discs from India made. And we're trying our best to do that. But, the truth is that it is going to move

27:36entirely digital, right? And the few people that continue to use discs, like Sony will figure out something, right? They will figure out for you to use an external disc provider. But, I think everything is going to move digital. But, there is an opinion I have, which is one step further, which is I think

27:49that it is going to move to streaming. Now, we've already seen some tests. Xbox has done some tests. And I By the way, I've I've been saying this for almost a year and a half, right? If you watch the first cut I did, I spoke a lot about the streaming idea. [music] And I think gaming is going to move exactly where Netflix has moved. It's the exact same equivalent. If you look at Blockbuster movies, there was a time when movies cost $100 million, $200 million, $300 to make. And we still have some of those movies. And they have a theatrical release. You pay separately for them. But, most movies are Netflix

28:14originals or smaller projects that are then licensed or sold to a Netflix or an Amazon Prime or an Apple TV. There are four or five, you know, big OTT platforms, and there's like four or five more in India that are smaller ones, even the US, Hulu and stuff. There are There are more than what most people are aware of. And they will commission shows with their budgets, and they make money on subscription. And I think it is the best model for the consumer to have one subscription you pay, much cheaper than what, you know, it would cost you. The cost of game development falls because of the new better tooling, because of

28:45artificial intelligence over time. And I think you'd pay one subscription to play many games, [music] and the payout to games will be based on retention, just like YouTube. So, if you play a game and you only stay for a few seconds, the platform is going to say, "Well, you didn't really contribute to retention." But, if someone's made an account just to play your game or come in for that game or has high retention on your game, then you get better payouts, just like YouTube. So, we're going to see a YouTube and a Netflix equivalent of gaming very soon. Xbox is trying to be that. Streaming now is very viable. This

29:14isn't Google Stadia from 5 or 10 years ago. I think bandwidths are good enough. I've been playing a lot of streamed games every day, and I'm in India, right? So, imagine that the rest of the world has even faster internet, and they're going to catch up pretty quickly. And the thing about this is this actually coincides with two other issues. The first one is physical devices are getting unaffordable because RAM prices are going up because these pieces of RAM are going to the data center. And what is going to happen is that some of these AI companies are going to do very well, but some of them

29:38are going to struggle and shut down, which leaves those data centers empty to stream games to you. And what will happen is you probably don't need to make the hardware purchase. Because the cool part about streaming is you need a screen, you need very minimal compute, and you need an internet connection, and a controller. And if you've seen some of the handhelds that are coming out, I really believe that the future of gaming is streaming. Just like watching a movie, we entirely stream it. But remember going to the store, getting a DVD for a movie, putting it in, me and my family sitting together, watching it

30:04on a TV, and once it's over, you return it back, right? You rented it physically, but now you're going to pay one subscription to this online store and be able to rent any game at any time digitally. So, I think it's going to move streaming, and I do believe like YouTube it will also move ad supported. In fact, Xbox did a test recently. I've spoken about this before Xbox ran this test, but Xbox did a test recently, which is that you're allowed to play for an hour, and then there are a couple of ads, and then there's more. Now, gamers instinctively are very angry with this,

30:29but I think in general this will bring a lot more people into gaming. This is right for the consumer because you have so many more options to choose from. You don't need to buy a game and feel like you've invested so much money and then not like the game cuz you're paying one subscription to have access to many more and the best titles will still have collector's editions. Like if you really want to buy a game, you should and will be able to buy them stand-alone. Like even today, if you really want to buy a movie, you can still buy some of these movies stand-alone. They of course cost

30:54a lot more than streaming the movie, but you can still own a copy if you want to. And in my opinion, the only reason it's happened to movies but not to games is simply because streaming has been impossible because you also need to provide an input into the game and that latency is going to cause issues, but we're nearing the edge of technology where that is almost solved, right? So this is going to be the future whether we like it or not. As a game dev, I have very mixed opinions about this. I want to have a collector's edition of my game. I want it to be physical. I have,

31:20you know, in my mother's house along with my Pokémon and Yu-Gi-Oh! Master Cards, I have a bunch of games that I've collected over the years, discs. I feel sad that that era is going away. I feel nostalgic about it. But at the same time, I know that fighting technology in the future is futile. We can shout about it, we can complain, the platforms are going to do what they want anyway. And ultimately, whatever we say, it is the consumers that vote. And as a game dev, to respect that. The discs are special to me, right? The ones I have at home, they're special to me because I lived a

31:44few years of my life in them in some ways. Like I have fandom towards some of those games because those are the only games I had back then. I even have fandoms to some demos. Like I remember playing this very early game called Soldier of Fortune and I only had the demo. The disc was only a demo. Uh we had got it with some magazine or something like that, but it was special to me because I replayed it so many times. I didn't have anything else. In some ways today, we're spoiled for choice. We have so many, so we don't care about any specific game or any

32:08specific fandom. But I do believe that it's the job of the authors of the game, the game developers, to build that fandom, to really fight against all the other things that are grabbing attention and build something that can immerse a player for X number of hours. We're trying our best to do that. There are many game developers across the world

32:23that are trying their best to do that. So I think a platform like this is going to have, just like YouTube came out and showed you so much more talent. We had the Curry Bakers making Obsession and stuff like that, so much more talent making crazy things. I think a YouTubization of games will actually bring a new era with this absolutely new

32:39kind of talent making the school stuff. But I do believe that now if gameplay is easier to build, then people will actually spend a lot of time building the cohesive story, the world. Like there's a lot more than just, you know, wipe coding a game, but that makes the job easier. Which means that there can be smaller teams building amazing games leading to the YouTube-ization of games,

32:56which allows everybody to participate. The difference between the old era of Bollywood and Hollywood and movies and now is that anybody can contribute. Anyone can go out and make a YouTube video. Your opinions, your ideas, expression, all of it, you have a platform for it, right? That's why I'm so so respectful of YouTube, right? Like who would have given me the opportunity to make a video this long and put it on a movie screen? They wouldn't have ever given me the opportunity. Which TV show would have allowed me to do it? But YouTube allowed me to do it. It gave me a space. So I believe that even if you

33:22have some crazy ideas, you'll find a niche audience. You'll be able to make those games, but the platform for streaming games is to come. We don't know what's going to be called. We don't know who's going to win it, but there's going to be a war, right? And that war is good for consumers cuz as they compete, the prices are going to be

33:37lower for us. That's it for me. Bye.