Transcription
One of my friends is building an AI startup, and he just sent me this. This is a graph that shows that they've grown by 260% in a day.
And we live in this weird time where the AI revolution is happening. And there is a chance that large companies like Google, Microsoft, and Amazon are going to take all of the market, cuz they're shipping so fast. They're able to create very powerful tools, but at the same time, they're not tailored to solve niche problems. And this is where you can make your millions.
The thing is, 2025 is the best time to be a solopreneur because you can code anything with lovable replet or cursor. You can design everything with Canva or other AI tools. You can market everything with Poppy AI—that's going to generate perfect scripts for you. So this is the time for someone who's been dreaming of building something to actually execute on that. I've seen a lot of people go from idea to an MVP in 24 hours. An MVP is a minimum viable product, something that you can show to a user and test it.
Like everything is speeding up, and the next couple years are going to be perfect for launching your own business. Maybe less. I don't know. Like I don't know how much time we have. But in this video, I want to give you some ideas of what you can build. And these ideas are based on what I see in the market as a creator, an entrepreneur who lives in Silicon Valley, but also on what Y Combinator is looking for. And Y Combinator, this accelerator that helps create billiondollar companies, and they always see what's going on in the market, and they're looking for very specific things that I'm going to name in this video. And I will also name some tools that you can use to build an AI company just by yourself, with no experience but a crazy idea and a lot of enthusiasm. So this is the time for people like us to build something and to bring it to the world.
You know, we used to talk a lot about content being commoditized and everyone being able to create content. I feel like we're entering—and we've already entered—this time where product is commoditized. Anyone can build a product now. So, it's your time. Let's dive deeper into business ideas in the age of AI.
Let me start with some numbers. Recent estimates suggest that the generative AI market could reach between 200 billion and 1.3 trillion by 2030, with growth rates of 25%. And it's always a good idea to be building a product in a growing market because this wave is just going to take you and push you forward.
Okay, I'm going to start with something that I'm building myself for my company, and it's called a content factory. As I mentioned, content has become commoditized, and both sides of the game have expanded. We are consuming a lot of content. We're also creating a lot of content. In order to stay visible in this new market, you need to be everywhere with your content. So, if you're a company like take Alo Yoga, they don't just have one Instagram profile. They have multiple, and uh they have uh an account for each specific type of content that they're creating. As a creator, I have to be on LinkedIn, Twitter, newsletter, reals, uh picture posts, YouTube podcasts, YouTube videos, clips. There are so many things I need to do. And in order to do this without hiring a lot of people, I need to build a content factory. And uh Gary Vee did it back in 2019, but he had hundreds of employees. I don't have this luxury of hiring as many people as I want, but now I can hire as many AI tools that I want. And businesses that help companies and creators build those content factories are in great demand. And there are some certain tools that you can already start using: of course, chat GPT for text, 11 Labs for voice. There are more targeted marketing tools like Poppy AI that lets you upload videos that you like from different platforms and upload LinkedIn posts, chat with it, and ask to create a script based on what you've uploaded. I think it's a genius tool. There is inv video that lets you create videos. But basically, even my company is looking for this all-in-one solution.
And what's really exciting when I noticed—and I'm talking to a lot of companies and creators—a lot of them are looking for people with no prior experience. We're in a situation last week where we hired a YouTube producer, and she—she has a hu huge background in producing videos—because everything changes every single week in our company. Like we wrote a script last week, but this week I'm looking at it, a lot of things has changed already. And uh I'm like, "We need to replace this, this, and that." She's like, "Oh, you know, in our previous company, every process was described. Everything was in order. We never rewrote the scripts, etc." I like, "I know. Unfortunately, like I can't do that in my company. We need to move really fast, and processes have to change every single week." So, we decided that we're going to start hiring more people who maybe have no or minimum experience, but we're looking for entrepreneurial-minded producers who will be open to testing new tools, open to rewriting scripts, open to changing strategies very often because this is how it works these days. Nothing is guaranteed for even like a couple months. You always have to be able to identify what's trending. And so, by the way, if you're that person, we're also looking for someone who is uh studying computer science and who's looking for an internship this summer. Uh we're looking for someone who can help us integrate more AI tools in our work and uh help us create more AI content. So, if you're that person, I'm going to leave a link below. Please apply.
And this is your time, guys. This is your time. If you're curious, if you're excited about AI, if you are willing to experiment, this is where you can build a business. And uh building a content factory, like offering that service is a huge advantage for you to make money this summer or in the next couple years. And it's always a good idea to build a product on a growing market because it's just going to take you and pull you forward. Because when the market is exploding, you don't have to push as hard. The momentum's already there. You just need to catch the wave early and hold on tight.
But what if you want to zoom out for a second and see where the wave is strongest right now? Want to know where startups are growing fastest? How AI is reshaping go-to-market strategies and which company might hit unicorn status with just one employee? HubSpot for startups just released the 2025 hypergrowth startup index, and it's incredible. Think 100 of the fastest-growing VC-backed startups around the world, all broken down by growth, team size, industry, and strategy. We're talking big shifts like Perplexity jumping to nearly 10 billion dollars. I actually had an interview with Aravent 2 years ago when they weren't on that level of hype, and uh the rise of new global hubs like Singapore and Bengaluru. If you're building an AI, this report is gold. It breaks down how investors are moving, how AI-native business models are evolving, and why small, smart teams are scaling faster than ever. My favorite takeaway is that customer support isn't just support anymore. The fastest-growing startups treat it like a secret weapon, turning users into advocates and advocates into growth. As always, the link is in the description, and it's absolutely free. And honestly, if you're serious about building something in this space, don't just keep up, leap ahead. Thanks again, HubSpot, for sponsoring this video.
The second cool idea is AI agents for small businesses that work 24/7. The emergence of autonomous AI agents is the biggest event in the industry in 2025. We're no longer talking about simple chatbots that just feed your replies. Now we have fully fledged virtual employees who can think for themselves, make decisions, and even learn from their mistakes. And the gap in the market is that everybody needs an AI agent, but it's kind of hard to create it now. You need to be more technical. You need to understand what's going on. But there are already very cool examples of AI agents. Uh there's a UK startup called Garfield AI that drafts legal letters for just two pounds without any lawyers. And there are emerging AI agents in every niche. There's Logistify AI which turns a voice message into a warehouse order in just 3 seconds. So the best way to enter this field is by creating a highly specialized AI agent that solves a very specific problem that you're aware of. Maybe you're solving your own problem as a specialist. For example, handling online orders for restaurants or finding gigs for freelancers, negotiating price, calling a hospital and asking for a discount because of this outrageous bill that uh doesn't really match your expectations. And the best part, you don't need to code with these. You can launch these projects uh on no-code platforms like voiceflow, cursor, replet, lovable. I'm currently playing with all of them. And it's amazing, guys, how you can just build with a voice prompt. Of course, it needs some time, and that's why there is this gap in the market where you can start building and deploying your AI agents for other companies.
Now, let's talk about how to make money selling prompts. And this is an easier idea. If you're watching this video and you're like, "Marina again, super complicated Silicon Valley, give me easy money," easy money for you. You write those little instructions that make AI do magic. And platforms like Prompt Base are full of people trying to buy high-quality prompts for like $15 each. What's the secret here? You understand the niche, and you make sure it's in high demand. For example, cyberpunk retro logos are super popular among startups or celebrity portraits for NFT collections. And I think there's an opportunity in creating bundles. For example, 50 fashion design prompts, and you sell it for $199. Top sellers on Prompt Base make $200 to $500 a month just by creating and selling prompts.
But if you're like, "Marina, okay, selling prompts, perfect, not Silicon Valley style," let's dig deeper into ideas that you can start small, but also they could lead you to actually building a billiondollar company. So these ideas are actually something that Y Combinator is looking for. So if you're going to start building something that I uh mentioned, please go ahead and apply to their accelerator. I think they're amazing. We tried several times. We never got accepted with our idea. Maybe it's coming in the future, but for now, oh, and something else I wanted to tell you, by the way, regarding AI agents, I recently joined a fund, uh where I will be investing together with the co-founder of Honest, uh the company that he did with Jessica Alba, uh the ex-CMO of uh Bumbo. She's also co-founder of LTK. And uh we're investing in products that are uh co-created with uh creators, with con creators because the game these days is not only to create a great product. Uh you need to find distribution, and that's why we're talking uh content factories. That's why we're talking AI content, and there is this AI agent that somebody built. It's like an AI agent that generates uh brand deals for you. So a lot of people are working on these AI agents. Anyways, uh there is another opportunity, uh something I feel like my fund would invest in. It's called AI tutors, and uh Reed Hoffman recently mentioned that they did an experiment with one of the schools where students used a personalized AI coach, and they did so much better on tests, and I feel like AI is overtaking lang the language learning industry. I feel like uh standardized tests may be a little bit harder because you still need the support of a real person. But when you're learning a language, you can just launch an app or a chatbot that's going to help people do that. And this is what we're experimenting with at LinguaTrip. And uh we released several apps in the past few months just because it's so much easier. A couple of years ago, it would take us a year to build an app. Now, we're able to do that in weeks. And uh one of the apps that we released is now number three in terms of how much revenue it generates. And this AI market—education AI market—is going to reach 24 billion by 2034. Profitability here is strong. You can charge subscriptions, you can work with schools, and uh your operating costs stay really low. I think that these educational chatbots are the future of um the course market. You know, when everyone was launching courses, now everyone will be launching their own app, teaching people their craft. And if you've been doing courses, please try building an AI bot yourself. Again, use one of the apps that I mentioned that allow you to code uh without knowing how to code.
The next big niche is voice AI agents. So, there are billions of business calls that happen every year. And now advanced voice bots can handle customer support, bookings, troubleshooting, calling that hospital and telling like, "Hey, how dare you charge $1,000 for a band-aid?" Or, "I think I was charged several hundred for ibuprofen." That's—that's your startup idea: build an AI agent that's going to go call a hospital about another outrageous bill. This is a relatively easy space to enter since you can build on existing platforms like 11 Labs. And the market for voice AI is set to explode, growing at about 35% a year and hitting nearly 48 billion by 2034. I feel like AI agents are just going to talk on our behalf, and uh this is your time to build an AI agent. And the profit potential here is huge. Companies can cut their call center costs by half or more, and customers get instant service.
Now, the next business idea I'd love to invest in—um, as a venture partner, but I'm—I'm super excited about AI in personal finance. Imagine having a personal AI assistant that knows what's going on with your apartment in Hawaii. Is it yielding something that you've projected or not? What's going on with your crypto? Is it a good time to sell? Maybe it's a good time to rebalance your portfolio. Maybe let's do some tax-loss harvesting. Maybe, Marina, it's time to think about buying a house instead of renting. Like all of these things. I do this all by myself manually. Yes, I use some AI tools to analyze expenses and do my bookkeeping. But imagine an assistant that tracks your spending 24/7, flags unnecessary expenses, suggests smart investment opportunities, and can even trade stocks automatically. Today, there are several solutions on the market already, like Cleo or Plum, and they're already outperforming human experts in speed and accuracy, and they cost a fraction of the price, often around $10 to $15 a month. The personal finance AI market is growing steadily, and the broader AI-powered fintech market is expected to hit 130 billion by 2034. When the market is that huge, we're going to have multiple players coexisting on the market, serving different target audiences.
Let's move up in complexity, and uh let's talk about AI for science. AI is now accelerating discoveries in chemistry, biology, and material science. What used to take years in the lab can now happen in weeks. And we've seen DeepMind's AlphaFold that predicts the 3D structure of proteins, saving years of lab work. And uh if you're doing your PhD, I want you to think about your PhD as—there's also my chance in building my own AI agent that is going to come up with hypotheses, test it, and uh suggest some tweaks to it. Uh basically your own AI researcher because the potential benefit here is massive. We're going to speed up research, and we can find solutions to problems that would otherwise take years and years.
Next is something really exciting, and it's robotic software. So, we recently had a dog—a robot dog—visit our house. And um when my daughters—Lily, who was 3 years old—she was just fascinated. For the next two days, she kept asking me about robots. She's like, "Mom, when we're going to have a robot in our house, what would it be able to do? Will it give me candy? Will it switch on cartoons for me?" Um, and uh we're talking about how it's going to wash the dishes, how when they are going to grow up, they're going to have a robot in their house taking care of uh laundry and dishes and cooking. And the way I see it is that robot is basically your iPhone, and then independent developers are going to be able to build apps for that robot. For example, I would need a robot to be able to set up a studio for me uh and watch the camera cuz I just recorded the whole video without my camera being switched on. So, I need a robot to make sure it doesn't happen again. So, I will be able to write an app for a robot to do something that I need. Maybe with my house specifics, maybe specifically in California because it's hot or specifically for St. Petersburg because it gets really cold. So, for now, the robotics software is mostly for B2B like delivery bots, warehouse robots, um self-driving cars, but the market is already worth over 24 billion and is growing at more than 20% a year. Profitability is high, especially in B2B sectors, and I feel like it's a very, very interesting niche that's just going to grow and grow. I already see Tesla bots delivered to a lot of content creators. I'm like, this is interesting, and non-tech content creators were seeking production with a Tesla bot. Robots are fascinating.
Now we get to full-stack AI startups. This means building entire companies like a law firm or a clinic where every employee is an AI agent or a robot. Of course, it's a bold move. It requires deep industry knowledge, but it's also not like a $100,000 company. It's potentially a billiondollar company if you have the ability to build complex systems. Full-stack AI businesses could be worth over 200 billion by 2033, with some of the highest profit margins in tech.
So as you see, guys, we are in the market where building products is being commoditized, and the way I am raising my kids is that I want them to be generalists and understand what's going on in the world—like, "Oh, this is happening in this industry, this is happening in that"—but also being specialized in one subject so that they can grab ideas from different sectors and implement them in what they're doing. And as an entrepreneur who started her first company at the age of 21, solving my own problem, um I would give this advice to everyone who's starting their business these days: If that's your first business, solve your own problem first because you become your own customer. Meaning that you're going to apply higher standards to your product. You're going to know exactly what you need from it. Uh and uh it's going to be so much easier for you to start that way. Later on, when you sell or exit this company or something happens to it and you stop working on it, you will be doing another process where you analyze the market and find a gap. But I feel like as someone who's just starting out, it's just great to solve something for yourself and build something for yourself using all of the tools that I mentioned.
Thank you so much, guys, for watching this video up to the very end. Let's stay excited about this AI revolution. Let's see it as a way of becoming a superhuman who can execute on ideas within 24 hours and test them and uh make the world a better place. Thank you so much for watching, and I'll see you very soon on this channel. Bye.