# Sam Altman On Miyazaki’s thoughts on art, Design Jobs, Indian AI, Is Prompt Engineering A Job? ## Метаданные - **Канал:** Varun Mayya - **YouTube:** https://www.youtube.com/watch?v=xFvlUVkMPJY - **Дата:** 06.04.2025 - **Длительность:** 21:26 - **Просмотры:** 440,587 ## Описание An incredible conversation with Sam Altman! In this episode, we dive deep into India's incredible AI adoption, creative ways people are using image generation models, and the future of jobs in the age of AI. Sam shares powerful insights on career evolution, AI's surprising impacts, and the exciting path ahead. If you are a brand looking for a collaboration, please email us at teamvarunmayya@gmail.com. 00:00 – Introduction 00:10 – Sam Altman on New Image Generation Tools 01:47 – India's Rapid AI Adoption & Usage 02:54 – AI's Impact on Creative Jobs 04:04 – New AI Jobs & Prompt Engineering 07:37 – AI & the Democratization of Creativity 11:23 – Are AI Startups Just Wrappers? 14:31 – Will AI Completely Automate Coding? 15:44 – AI’s Potential Impact on Indian IT Jobs 17:36 – The Future of AI Agents 18:53 – Should You Leave a Company That's Slow on AI? 19:43 – Sam’s Counterintuitive Insight on AI 20:52 – Conclusion ## Содержание ### [0:00](https://www.youtube.com/watch?v=xFvlUVkMPJY) Introduction Ladies and gentlemen, I'm with Sam Alman. And Sam, uh, I've seen a bunch of your tweets over the last few days. They're called posts now. They're called X's now. I don't know what they're called anymore. called. I still call them tweets. You've been tweeting a lot about ### [0:10](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=10s) Sam Altman on New Image Generation Tools India. You've also the new image gen. Firstly, I want to say I've had early access to the photo image gen thing. It's been great. Everything from mood boarding, making images, UI, it's just been phenomenal. In fact, it's also very smart. So we've been we have this very clever use case where we're generating the front view of a 3D model and then asking it for the left side view, the right side view, the view behind and then we're throwing that into you know another model that generates the mesh and then reproing these textures on top and then also asking it to generate height maps which is just the insanity of this existing where it didn't exist a couple of months ago is just it it's it blows my mind. What's the response been like? Any numbers for us? I don't know if I can share an exact number but like way more images than I thought the internet had uh like appetite to do like the um we will cross a billion total images in the notistant future. I can say that. Oh man, that that's a huge number. And what's the response like any creative use cases that you've seen on images so far apart from you know animating yourself? Yeah, I mean the thing that's been exciting to me is just the breadth of creative use cases. Obviously turning yourself into anime was the one that kind of first lit up on the internet, but it's been amazing to see the sort of creative expression, people making just really new kinds of images I had never thought about before. Um there's obviously a lot of great commercial use cases too where people are, you know, using this in their small business for new logos or graphic design or whatever. But really the creative spirit of people just doing things that, you know, images, kinds of images that I've never seen exist before, that's been to me quite exciting. And ### [1:47](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=107s) India's Rapid AI Adoption & Usage has India contributed? I mean I think India has been a big user of this from your tweets at least. India I believe is our fastest growing market now. Um India's been very special to us for a long time. Um India was one of the first markets outside the US that really jumped on AI in a huge way and since this moment has happened it is now our I'm pretty sure it's our fastest growing market. Wow. That's that's so it's been very cool to see. That's crazy too. And have you seen Indians also use things outside of imagen? Have you seen them use deep research? Are they a paying market? Uh I was in India a couple of months ago and the that was obviously before image genen and you know it was like non-stop stories from people about what they were using chatbt for um to do not catch 40. Um but it was really it's like quite incredible the degree to which it has permeated Indian society for all sorts of things. You know certainly we get feedback that our price point is too high for mass adoption in the Indian market and we're looking at things to do there. Unfortunately, our compute costs are still just quite high. But we are hard at work on more efficient models and I'm optimistic we'll be able to bring cost down over time. Amazing. And ### [2:54](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=174s) AI's Impact on Creative Jobs I want to touch upon the other side of this, right, which is the UI and design jobs for a second. You know, we thought when these tools come out came out a year ago that well, we're going to have, you know, these tools do most of the work for us. But, you know, in our own company, we have a bunch of people who are video editors. We have people who are designers and it turns out we still need a person to run the tool because humans still need accountability someone maintaining context and it's almost like the manager of the tool in some senses right uh is this how jobs are going to evolve is have you seen something similar or do you have a path for how jobs will evolve now that these sort of tools are here I think it's going to be uneven and it'll be different for different kinds of jobs there will be some jobs that are that totally go away where the AI just does them end to And mostly I think it'll be a case of a new tool where people are just much more productive and can do work at a higher quality. So there will still be people that are in charge of you know making a website look great. But the expectations of how good a website looks just go up a lot and we all get better stuff. Uh in fact maybe we get more maybe there's more demand for that in a world like that that I think can cut many different ways. Um ### [4:04](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=244s) New AI Jobs & Prompt Engineering and then there will be totally new classes of jobs that we just haven't seen before at all. Yeah. Do you have examples so far? There are some kinds of customer support where I think an AI can just do that end to end and that'd be in the first category. In the second, since we're on this example of graphic designers, you know, taste still really matters and like the you know, when we went from graphic designers had to do everything with like uh paper and pencil or you know, pens and paper, whatever you want to call it, to having computer tools, they were able to do more and do better, but we still needed a lot of good graphic design in the world. Um, we can now have a new tool, but and no doubt it's going to change the workflow of a graphic designer and some things that some people were paid for, they won't be paid for those tasks. But what it appears to me will happen is you will still need some sort of job, which is, you know, make this website look beautiful. Um, and maybe there are fewer people that do that, but they make much more money. Maybe there are more people that do it because we just have a, you know, explosion of how many websites we can get. Like may maybe it turns out there was way more demand for graphic design in the world than we could sort of like afford to fill. I think we just don't know on that yet. Um, and then there's other and then in terms of like entirely new jobs, you know, the first example we saw of that was prompt engineer. Before AI came along, there was no such thing as a prompt engineer. Not only that, there was no it would have been hard to conceptualize that there would be such a job. Um there's many more, but that's the example I give is the sort of near and dear to my heart one. And it's so funny when whenever some of these new jobs come in, they're always ridiculed at first, right? Like even building rappers, and we'll come to that in a bit, but even with prompt engineer, a lot of people still don't take it seriously. Well, all you're going to do is write English. Do you have a defense for that? Do you feel there's some way to, you know, make the people who are now going to be prompt engineers feel a little more, you know, like they're not losing status for doing it? When I was young, um, I knew I wanted to go study computer programming. And the adults in my life would say, "Oh, that's like that's a hobby job. These computers aren't going to go anywhere. You need to go be a doctor or a lawyer. Like, you need to get a real stable job. " You know, this is like a this is a dangerous path to go down. Um, and I thought I understood something that they didn't. And I think I really encourage people to have their own conviction on what things are going to be like. And just because something is not a historically valuable or high status job doesn't mean it won't be in the future. Um this is like the way of technological progress. You know for people that are watching there's a post by Sam Alman on this which is the strength to be misunderstood. I love that blog post by the way. So thank you so much for that. You wrote it many years ago. I think it was very insightful when I decided to become a content creator and it was a low status job back then. I was going to say about your job, you know, I remember, you know, like 10 years ago or something. Uh, I would like I would meet people I was, you know, I would go talk about doing startups or whatever and I would like go meet young people who wanted to start startups. And I always kind of like asked people what they wanted to be or what they thought the interesting industries were because I think if you kind of like ask young people that you get a perspective that is interesting and different and usually right. And every not everybody many people said they wanted to be a YouTuber at the time. I was like what? But I it hit me then that I'm like ah I cannot dismiss this like people dismissed computer programmer when I was young. Yeah. I ### [7:37](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=457s) AI & the Democratization of Creativity want to touch upon something else that the internet's been sort of debating over the last few days which is you know there's this video of Miyazaki. So of course it's an out of context video where he's looking at a model from you know many years ago something some creature thrashing around and he said that well you know I don't like this. Essentially that was the you know the summary of what he said. you know, as an artist who's been making videos for n 9 years, I feel some of the craft and the, you know, the effort I'd have to put in to make an image or even to make a video has dropped over the years. And it is a little bit upsetting. It feels like it's a little bit of a bummer as an artist, right? But as an entrepreneur, as somebody now that has a team to manage and worry about their salaries and payroll and whatnot, it is very good that this technology exists because it helps me go a little bit faster, right? It helps me, you know, make sure that I'm dealing with a lot less stress. Do you have a view on um, you know, how the internet should think about this evolving nature of art? I think what it was like to try to um do what you do 30 years ago. It was impossible or very hard. I mean, you'd need like a camcorder and like VHS tapes and then you'd have to like edit it in some complicated way. And then even if you did that, you would have to like distribute tapes because there was no internet and no YouTube or there wasn't internet but it wasn't very good. And now like anybody on earth can get a smartphone and record themselves and put it up and if they have something interesting to say they get it out there and the world benefits from that. Um I think the democratization of creating content has been a big netwin for society. It has not been a complete win. There are negative things about it for sure and certainly it did something about the art form but I think on the whole it's been a win and this is something that I very deeply believe about the power of technology which is giving everyone more tools making things easier lowering the barriers to entry um does significantly increase the number of people that can contribute to society and we all benefit from that overall. Doesn't mean that it doesn't like cause some job loss and some people who had a sort of differential ability to do something now have a lot more competition, but overall I think it's a real benefit to society. Um my own experience of this was watching the barrier to entry to starting a company really change. Um, this happened when it got cheaper to start a company because of a whole bunch of things, but AWS was one of the biggest and most important moments. Access to the internet in general was another big one. And there was a shift that Y Combinator I think really helped drive or popularize whatever you want to call it. uh starting in 2005 where young technical founders could all of a sudden much more easily start companies and that led to an explosion of great new products and technologies that the world got to benefit from. Now some of the people that those companies competed with probably wish that hadn't happened but as a citizen of the world there was a lot more stuff. OpenAI itself, I think, was an example of a company that only got to happen because the barriers to entry to a bunch of different pieces of technology stack got significantly low and this ragtag bunch of us were able to just do something we had no right to do. Yeah, I think I agree with you. I think especially after Stripe and in India after UPI, there's just been an explosion of new companies, new types of things that were just not possible because the amount of red tape that it required to even get started. I assume now it's sort of become the same in design. In fact, on ### [11:23](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=683s) Are AI Startups Just Wrappers? the note of YC, right, I actually, you know, I want to ask you to put on your 20-year-old when you were a 20-year-old hat of, you know, going to YC again. If you knew everything you knew right now and the debate was, you should build a rapper or no, you shouldn't build a rapper, OpenAI would probably kill your startup or on the other side, foundational models are worthless. Rappers are the only ones that are making money and able to raise money and whatnot. Um, what would you do knowing everything you knew, but also with the ability to shut out the noise? I think the whole like oh your startup is just a rapper or oh like people used to say that about other categories too. You know there was a time when people say well you're just a rapper on AWS probably. Um the people are building absolutely incredible new companies based off of AI. Most of them will fail or not do that well, but some of them will find a really enduring business and generate a ton of value. And that's always the case. That's just how the world work. Like, you know, most startups fail and some really succeed. And it's easy to sound smart and to just dismiss everything as a rapper or some thin thing that's not going to, you know, create much value. And I remember when people used to say very similar things about open AAI. Yeah. What was it? What was the they used to say? You're just, you know, you're just like a little research lab and Google's going to crush you and like you're just putting this like thin layer on or Nvidia is going to get all the value or you're going to like you're doing this like little thing and it's impossible to compete with the big tech companies. And it was not long ago they were saying this. This is a couple years ago. So I don't know. I it sounds very cool and sophisticated to always say why things aren't going to work and you will be right most of the time. Um and kind of the most important lesson I think you can learn as an entrepreneur is that or an investor is that is not how great companies get built or invested in. So I think if I were 20 in NYC right now and someone was calling my company a GPT rapper, I'd be like, "Ah, I'm I must be, you know, in a promising direction. " Do you have any specific ideas that you would be working on if you were 20? But you were not an open AI, you didn't have the resources. You had 500K. As a matter of principle, I try to never answer that question because I the thing I care about most is that people don't start someone else's company. Like you the ideas that I have for companies are probably mid at best. And the idea that some 20-year-old doing YC right now that has a company that has an idea for a company that sounds totally crazy and bad to me, that's probably the next open AI. And I never want to discourage anyone from that by saying some like safe middle of the road like out of touch idea. So go do your crazy idea. It's my advice. That makes sense. It's almost like you need to do what you're passionate about. Otherwise, you never end up building a Snapchat because everyone would think it was a bad idea on day one. Um, that's amazing, Sam. You know, we did a we did ### [14:31](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=871s) Will AI Completely Automate Coding? amazing, Sam. You know, we did a podcast with Kevin, uh, your CPO the other day, and he gave me a timeline for when code is going to be automated. He said 2025. Anthropic said 2027. He said, "No, we can do better. We'll do 2025 or maybe 2026. " Um, now that we know everything we know, do you have a timeline for when, you know, a lot of writing code becomes English? or is that not going to happen? I think it's the degree of automation that matters to get to 100 truly 100% automation. You know, you can make a complex thing and never touch code. That that's one thing, but I'm less interested in that question than when a coder becomes 10 times more productive. And I think that could happen this year, next year, something like that. So, a coder is 10x more productive. I think a coder is already several times more productive. I think there's already like at least in you know the way people around me have started working the good ones who've adopted the yeah people maybe say 2x 3x but I haven't heard anyone I haven't heard many people I have not heard a consensus for 10x yet but you think that by next year it's going to be a 10x step function look predictions especially predictions with the timeline attached are always difficult but I think we can aim at that yeah interesting you know there's there is a ### [15:44](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=944s) AI’s Potential Impact on Indian IT Jobs negative side to this right which is that you know Daniel Grath had this one article where he said that $250 billion of India's exports are actually IT services and if you think from that angle that could just be GBD4 tokens right or GDP5 tokens at some point for India that $250 billion is very important and not all of those engineers are building are elite engineers who will benefit from this technology and go 10x right some of them are still building simple crab crowd apps and there is competition from the 10x engineer maybe not from AI itself but an engineer using the AI and sort of outpacing them and It's they're able to do this at much cheaper, right? Because the tokens are of course cheaper. Do you have something that India can do uh to actually make sure we absorb some of that? We continue to be relevant there. If there's one area that I think this is an important general question, but if there's one area where I think that the world just has so much more demand than we can currently supply, it's for codewritten. My belief is that the world just wants way more software and is about to get it. And if you give people these tools to write software much more effectively, I at least think for a while what will happen is someone who today can write code at some value will be able to write code or create software let's say at some much higher value. Now they have to pay the AI a lot so they won't capture all of that. Um and also the market price for a piece of code will go down. But I would bet this is an example. Everyone seems obsessed with Jevans paradox all of a sudden and I would bet this will be an example of it. So a lot more software being built. So this it might the cost might drop a certain amount but then also demand goes up by that amount. I think so. Interesting. That's very reassuring to hear actually. I don't think it's going to be like that. I don't want to like dodge the question because I don't think it'll be like that in every industry but code is one where I would guess it's like that. And do ### [17:36](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=1056s) The Future of AI Agents you see a lot of these I mean we've been seeing sort of early agents right? We've seen deep research. D research has been great but DP is also a passive agent. It's not taking my card and doing something with it. So there's low DP research is low risk. Um I've seen that you have been very bullish about agents. There are lots of businesses that have been very bullish about lots of other companies that have been bullish about agents. Um do you have something in mind for what agents will become over the end of this year or by next year and some use cases that you think we'll see on an everyday basis with agents? Just to stick with the code example because we're on that. I think this idea that you can like say to a coding agent um I would like this new feature built and it goes off and pings you back you know some period of time later and says okay here's the pull request ready to go that I think will feel very crazy to people there's other categories where it kind of goes off and does stuff for you on the internet operator points at that a little bit um but I would expect the coding one to be a big anything on a consumer use case like something like booking your food or something ordering a cab or something like that. That's the kind of thing that I very much expect to happen and that that'll feel like a very agentic task. ### [18:53](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=1133s) Should You Leave a Company That's Slow on AI? Let me ask you another question which is if you were an employee at a job but you were working at a legacy company let's say they were slow to adopt all this technology they're not even aware that you know maybe they've tried it in passing they've seen a GPD3 and they've become dismissive about and lots of legacy companies are like this if you were sitting in that company and you were unable to convince your company to use these new tools because of bureaucracy or whatever reason would you what would your career plan be then would you move to a faster moving company but also you know take on the risk of that company probably I would do that. Yeah. Both because I think companies that don't adopt these things will be in a tough place. Most I mean, there's some categories of companies that of course be fine, but also because I'd want to be responsible for my own career success and I'd want to make sure I was learning and you know, if I felt like I was not getting exposure to these tools, it would seem super broken. That makes sense. Okay, one last question and then ### [19:43](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=1183s) Sam’s Counterintuitive Insight on AI we should be good. Um, can you give me a counterintuitive insight you've had about AI? worked here for a very long time of your life and you've done a bunch of things. You've done YC where you've invested in companies. You've you've been around the valley for a very long period of time. What is something you thought about AI but you have a totally different experience now? I somehow thought it was going to be a bigger change to society than it's turned out to be. Um, you know, I I had this sort of like sci-fi like vision of what AI was going to be like. And maybe it'll still get there, but right now it's just this like crazy powerful tool. And life goes on mostly the same. And I'm actually very happy about that. I think that's good. I think I was naive before. I also think that would have been bad. But it's it it's very different than I kind of imagined it 10 years ago. But I think that's wonderful. Yeah, I think that's wonderful, too. I think you know having godlike technology being able to solve at some point medicine and help us do less of the work we hate but more enjoy I think is is net good and I hope that you know everyone benefits from it at some point. Um thank ### [20:52](https://www.youtube.com/watch?v=xFvlUVkMPJY&t=1252s) Conclusion you Sam. Uh this was a wonderful interview. Thank you so much for being support and doing this you know on such short notice and I hope uh I hope you know India continues to use this technology and I have no doubt I I what is happening with this technology in India I have uh I've never seen adoption like this of anything anywhere and it's very cool. Yeah, I really hope we do more of that and I hope you know we put it all over the timeline and you know you see some of the things we you created. Thank you. Awesome. Thanks Sam. Bye. --- *Источник: https://ekstraktznaniy.ru/video/12022*