Are We in an AI Bubble? (Sam Altman Warns YES + Your 2-Path Playbook)

0:00Are we in an AI bubble? Now, Sam Altman says we are. But today I wanted to investigate it. I wanted to make the case for if we are in an AI bubble, yes. And if we aren't in an AI bubble, no. And then I wanted to end this episode with what this all means. What does it mean if we're in an AI bubble? What does that mean for your startup? And if we aren't, what does that mean?

0:32So let's start with the case for we're in an AI bubble. Why would people say we are in an AI bubble? The first reason is there's just tons of hype around AI startups. When you scroll on X today, I know that you've seen some of these demo videos of AI startups. They're cinematic. They're well done. They probably cost$100 ,000 to make, and they're everywhere. It's standard. It's the equivalent of when I was coming up in startups, being on TechCrunch, you had to be on TechCrunch. That's the demo video today. People spending tons of money on demo videos. The other thing about tons of hype is people say that marketing is trumping the AI products, that there's a lot of products that aren't that good relative to their marketing.

1:29Now, I actually haven't used the Clueli product, but Clueli is like the poster child for the best marketing right now for AI startups. I saw this Instagram post today. Clueli engineer buys a$500 ,000 Ferrari. This is just marketing that gets people talking. Now, the question is, is the Cluely product as good as their marketing? Maybe it is. Maybe it's even better. But there are a bunch of products that have really good marketing that the product people try it and they're like, it's just not there. The second reason why people say that there's an AI bubble is that the valuations are getting insane.

2:14Seed round valuations, it's not uncommon for them to be 20 to 30 million dollars. When you think of that, think of a 20 to 30 million dollar building you could buy. That's the equivalent, right? Two or three people start a company and all of a sudden their business is worth 25 million dollars. That is a lot. You know, that is twice what it was five to 10 years ago. Now, the other thing is seed rounds that are in the$5 to$10 million range are pretty common now for a hot AI seed startup. My friend Riley Brown, friend of the pod, he just raised$9.2 million for his VibeCode seed round. When he just started the company, he raised$9.2 million.

3:03Of course, he's an incredible founder, but$9.2 million for a seed round, this is something that we haven't really seen much before. That is a tremendous amount of money, especially for a first-time founder. Yes, in the past, we've seen 9, 10, 12, 15 million dollar seed rounds, maybe for a second, third, fourth-time founder who's had a huge exit. But this really talented group of people, AI founders, raising 5, 10, 15, 20 million dollar seed rounds are getting quite common. The other reason why people say there might be an AI bubble is a lot of these AI startups have the same base model. So it's really all about distribution.

3:57It's almost as if the VCs don't really care that the underlying tech is the same. It's almost like they're not sober in that sense. So that's why some people say that there's a bubble because the VCs are just funding these thin mode companies. The last thing, one of the main reasons why people say it's an AI bubble is there is shaky unit economics with AI startups. So what they say is, I saw this tweet that summarizes it quite well. I got 2 million views. Why is AI a house of cards? Number one, you pay$200 a year for an AI app like Cursor. Cursor pays OpenAI$500 for API tokens,$300 of which is VC funding.

4:50Number three, OpenAI pays AWS$1 ,000 for compute,$500 of which is VC funding. Number four, AWS pays$10 ,000 for NVIDIA GPUs. Do you see the problem? Unless you as a user are miraculously comfortable paying$1 ,000 for an AI app, the only thing propping up AI is VC funding. No VC funding, the AI application layer is unprofitable, the LLM layer is unprofitable, the compute layer is unprofitable, and the GPU layer is unsustainable. Now, these are the main reasons why people say it might be an AI bubble. And I think when people see stories like icon.com raising$25 million and then spending$12 million on the domain, it just exacerbates it.

5:46So I can understand why people say it's an AI bubble. I'll let you, the listener, the watcher, decide for yourself. Is it an AI bubble? Yes or no? Let me know in the comment section. Sam Altman, the co-founder of OpenAI, just said that it is the era of the idea guy and he is not wrong. I think that right now is an incredible time to be building a startup. And if you listen to this podcast, chances are you think so, too. Now, I think that you can look at trends to basically figure out what are the startup ideas you should be building. So that's exactly why I built IdeaBrowser.com. Every single day, you're going to get a free startup idea in your inbox, and it's all backed by high quality data trends.

6:35How we do it, people always ask. We use AI agents to go and search. What are people looking for and what are they screaming for in terms of products that you should be building? And then we hand it on a silver platter for you to go check out. We do have a few paid plans that, you know, take it to the next level, give you more ideas, give you more AI agents and more almost like a chat GBT for ideas with it. But you can start for free, ideabrowser.com. And if you're listening to this, I highly recommend it. Let's move on to if it isn't an AI bubble. What are the arguments for it's not an AI bubble?

7:15The first one is real utility. So a lot of people are downloading apps, buying B2B software, and it's taking tasks that they would take hours to do into minutes. There's also revenue facing workflows like people are actually, you know, implementing this in their workflows. It's helping them save time and money. The greatest example, you know, one of the great examples of real utility is, you know, the CalAI guys who I think are doing$2 million plus MRR. In the past, you know, if you wanted, you know, I'm sure a lot of you know CalAI, it tracks your calories for food. Now you just take a photo with AI and it basically tells you how many calories.

8:01It obviously makes the workflow a lot simpler. So there's real utility there. There's real time savings. So how can it be a bubble? Costs are, number two, costs are falling. So now it's cheaper training and inference and there's open source pressure. So the cost of basically, you know, building and maintaining these apps is going down. And a lot of people actually don't know what inference is. So I actually wanted to take a quick side quest to explain what inference is. is. So inference in AI refers to the process of using a trained model to make predictions or generate outputs based on new input data.

8:44So simply put, training is when AI learns. Inference is when it uses what it learns. So the analogy, how do you think about it? Think of AI training as teaching a chef every recipe in the world. Inference is the moment you say, make me lasagna, and the chef whips it up on the spot. No more learning, just doing. So the cost of inference, which is really the foundation of AI, is going down and down and down. So, you know, when we're talking about the shaky unit economics, maybe there is shaky unit economics, but if the costs are going down, maybe it can get profitable. The adoption curve. So there's, you know, an increasing adoption in a lot of these AI apps.

9:33I think ChatGPT has 700 million active users. Co-pilots are being added in core apps every single day. And there's been a ton of multi-year enterprise deals. So a lot of the SaaS product, a lot of companies have signed deals with AI SaaS companies. It's not like they're going to pull it out of their workflows tomorrow, right? So how could there be a bubble if they've signed a five-year deal, a three-year deal, or two-year deal? There's also new moats that are existing. So there's proprietary data loops that are happening. There's workflow locked in. We kind of talked about that with the adoption curve.

10:20And there's durable CapEx, right? There's chips and data centers. There's power buildouts. These are all reasons why there might not be a AI bubble. So what does this mean for you? How can you play this AI wave? Well, if it is a bubble, let's talk about that first. You're going to want to focus on cash flow for services. If you believe that the bubble is going to burst, you're just going to want to have a low burn business that is making cash flow. The second thing is if you work at one of these AI startups, don't assume that your equity is worth millions of dollars. because a lot of funding is happening.

11:11And if the rug does get pulled, your company that you think is worth$500 million on paper just might not be worth$500 million. So this has happened time and time again and other booms, the dot-com boom, of course, the mobile boom, the cloud boom, the social boom, this happens. If you think it is a bubble, you also might want to think of having small diversified bets. You probably might not want to have all your eggs in one basket. And the last thing is, and I know people are going to say, I can't believe you're saying this, Greg, raise VC. If it is a bubble, you as a founder might want to raise VC.

11:56Why? Because you can get a lot of money right now. And if you can create a sustainable business, and if you understand that you are in a bubble, but you're going to raise this money and you're going to create a cash flowing profitable business, maybe you don't raise again. And it's sort of a one and done round. Or, you know, you just, you're very conscious about, you know, using that capital. You don't spend it all in one place. I know a lot of you, you know, you know, I love bootstrap businesses. Right now I'm building a bootstrap business and I'd love it so much. I'm actually having the most fun in my career doing it.

12:27But I will say like, if you believe that there's a bubble and you believe it's really raised money and you don't raise too much money, you raise the right amount, there might be a play for you there. If you think it isn't a bubble, you're just going to want to remember that distribution matters a lot. You're going to, you know, want to Yeah, basically focus on distribution. Focus on getting customers. Don't focus on getting every customer. Focus on the customers that fit your ICP and play the long game. You're going to want to expand, right? If you don't think it's a bubble, try to incubate a lot of things.

13:11Try to buy companies. If you think this is the long term, this is going to happen for a long time. You can take bets. You can take some of those bets if you believe in it. You're going to want to own the data, not the models. And you want to remember that the community is the moat. What is always true is no matter if you think it's a bubble or not, you're going to want to talk to users weekly. You're going to price for profit early. Retention over trials. Retention is everything. And then remember that there's always an arbitrage around CAC. Meaning, there's a little typo here, I'll fix that. Meaning, no matter if it is a bubble or it isn't a bubble, there is some way, some channel, some messaging, some creative that through paid ads, you can get a low enough CAC to get your business profitable customers.

14:13Now, the hard part is figuring out what channel that is. and what creative that is and what messaging it is and what positioning that is. But it's always going to be true. So I want to end with a few things. I want to talk about what happens in a bubble. Well, even in a bubble, it'll start with hype. And by this is someone I've lived through a few of these bubbles. It starts with hype. There's a crash. There's a shakeout. And there's real value. Even in the dot-com era, right? Everyone says, oh, the dot-com era, pets.com raised at this crazy valuation and it was nothing. Amazon came out of the dot-com era, which is like a trillion dollar company.

14:57There's going to be, if it is a bubble, there's still going to be a trillion dollar company that comes out of this era. And there are benefits to a bubble. when there is a bubble there's also a lot of interest there's just a lot of consumer interest so distribution tends to be easier during a bubble people hear about, let's just say if it is an AI bubble, people hear about AI hear about it and they're interested in it so distribution becomes easier easier to get customers that gives you data, that allows you to create a better product which helps with retention and creates this flywheel so I'm not gonna I'm not gonna tell you what I think I'm not gonna tell you if I think it's an AI bubble or not I I know I've got the smartest audience of startup ideas podcast you are all super smart I want to see in the comment section what you think this has been um hope this is I hope this got your creative juices flowing and I'll see you next time