The Deep View: Conversations

AI bubble talk is rearing its head again, but the context is very different from the conversations in late 2025. 

In this episode of The Deep View Conversations, we unpack the common arguments about an AI bubble and explain why reality naturally falls somewhere in between the doomsayers and AI absolutists. 

We look at AI's "Tinker Bell problem": the boom depends partly on people continuing to believe in AI's potential, even as public skepticism grows. Beneath that belief cushion, enterprise contracts drive most of AI labs' revenue, while strong hyperscaler earnings and compute shortages suggest durable demand is building. 

We debunk a viral claim that a $200 Claude subscription costs Anthropic $8,000 to serve. We also look at enterprises' push for more control, efficiency and measurable ROI, including one company's claim that some engineers' token use costs 1.5 times their compensation. 

Other topics include:
• Training, inference, API pricing and token economics
• Real value, snake oil and the hype cycle
• Why AI demand outruns compute supply
• Why the AI bubble may look more like bubble wrap
• Market rotation into energy and materials 

If you're trying to separate durable AI demand from hype and understand where a real correction could begin, then this conversation offers a framework for thinking about what may pop, what may deflate and what may keep growing. Keep in mind that this is industry analysis and not investor advice. 

Subscribe to Deep View Conversations for interviews with the leaders shaping the future of AI, business, and technology: tdv.transitor.fm 

And don't forget to sign up for The Deep View daily newsletter. We don’t just cover AI, we decode it. In a world flooded with hype, we deliver sharp, no-nonsense insights to keep you ahead of the curve and help you put AI to work every day: subscribe.thedeepview.com 

Creators and Guests

Host
Jason Hiner
Editor-in-Chief of The Deep View
Host
Nat Rubio-Licht
Senior Reporter at The Deep View

What is The Deep View: Conversations?

From frontier labs and enterprise platforms to emerging startups reshaping entire industries, The Deep View: Conversations podcast interviews the brightest minds and the most influential leaders in AI.

Jason Hiner: Alright, Nat. Let's talk about bubbles.

Nat Rubio-Licht: Let's talk about bubbles. A lot of the bubble talk has died down a lot lately, but I think that there are still a lot of misconceptions out there. There are still people that think that it is going to pop in a big, fiery explosion that's going to ruin everything. And then there are people that- This is all a mirage. Everything's fake. And then there are people that think that the bubble isn't real at all and that everything in AI is ironclad, which is simply not true.

Jason Hiner: We know. It's also-

Nat Rubio-Licht: The reality is somewhere in the middle, but I think it's important that we clear up some misconceptions.

Jason Hiner: Yeah. The thing is that there's a lot at stake. There's a lot of money that's been invested in AI. There's a lot of people investing their careers in it. There's a lot of companies investing huge amounts of budget in it. And then we know in terms of venture capital and all that, roughly, by some estimates, half of all of venture capital is going to AI. It's an incredible amount of money, like an unprecedented historic amount of resources betting on this thing. And so I think we know we have to believe there are pockets of this that are going to be inefficient, meaning that's probably where there are bubbles, right? And those bubbles are going to have to pop or deflate or whatever. But also, because people have sensed that, they're sort of looking for boogeymen, right? They're looking and saying this thing or that thing. It's understandable because we know that there's parts of this thing that are going to run out of energy. It's like the natural process. You let a thousand flowers bloom. Eventually, you're going to have to cut down some flowers.

Nat Rubio-Licht: Or some flowers are just going to wilt and die.

Jason Hiner: Or they're just going to wilt and die. They're going to die a natural death.

Nat Rubio-Licht: What are the metaphors? Can we use bubbles, flowers? Tinker Bell?

Jason Hiner: Tinker Bell.

Nat Rubio-Licht: Yes. Yes. So here's the Tinker Bell metaphor.

Jason Hiner: All right. Hit me.

Nat Rubio-Licht: In Tinker Bell, the way that children or whoever the characters in Tinker Bell are that I'm forgetting conveniently right now, the way to keep Tinker Bell alive is that you have to believe in fairies. And I think that that is an apt metaphor for where we're at with AI because we are sold on this very utopian vision of what AI has to offer that it can completely transform our society and the way that we work and the way that we live and our health care and our infrastructure and everything. But only if we believe in it, only if we allow it, only if everybody buys in. But I think that we're not going to get the buy-in that we are thinking or that at least that the labs are hoping. We're not getting it right now. The narrative among the public. 61% of people don't believe in—

Jason Hiner: Or have a negative view of AI from our data and from other people's data too. So we know that 61% of people in our audience and in the broader in the US have a negative view of AI. Yeah. So I mean, we talk about it a lot on The Deep View that AI is losing the narrative among the broader public.

Nat Rubio-Licht: Yeah.

Jason Hiner: But among the people who are investing in it, working in it, you know, basing their careers on it, man, they are all in. Like we see it when we go to conferences, other places, like they are so in like this is changing everything. This is changing the future. This is changing all the things. And it's that part. That's the part that you're getting at that like that's sort of lifting up the investment that people moving careers to get into it and all of that. And that is like, you know, this sort of upward force that's driving it. But if they stop believing, uh-oh.

Nat Rubio-Licht: Yes, exactly.

Jason Hiner: Uh-oh.

Nat Rubio-Licht: But I guess how much, this is why I think it's going to fall somewhere in the middle. How much belief needs to happen in order for it to like, who's who needs to buy in? Who needs to like, I know that because these, um, there's this misconception that these companies are making, they're not going to be able to make all the money that they're spending. And I do think that they have spent a lot of money. They're spending a lot of money in the build up, spending a lot of money on compute, they're spending a lot of money. When I say these companies, I'm talking about the big AI labs, spending a lot of money to serve up compute. Um, a lot of that is going to training, however, not to inference. Um, and a lot of that is also going to building out data centers. Um, so I think that costs are going to come down, but, um,

Jason Hiner: right.

Nat Rubio-Licht: And also the other, the other part of it is that most of their money isn't coming from the subscriptions. Most of their money is coming from enterprise contracts, correct?

Jason Hiner: The revenue,

Nat Rubio-Licht: the revenue, sorry.

Jason Hiner: Yeah, is coming from the enterprise. Yeah, like big time.

Nat Rubio-Licht: So are the enterprises the only ones that really need to buy in? And if so, why are they spending so much time trying to sway the public narrative to get consumers to get,

Jason Hiner: yeah. So yeah, because they need to get a lot of people to buy in because those people are employees. Those people are going to start companies. Those people are going to start startups. Those people are students who are eventually going to influence, you know, AI adoption in that, in their companies, in their future companies. So they need everybody because that's also going to sort of lift this general, you know, adoption from the enterprise. But I think make no mistake, at least for the, you know, the coming future, enterprise is going to continue to power, um, AI and AI adoption. And as enterprise spend goes, like so will AI adoption. I think that's why you've seen OpenAI pivot so hard toward enterprise in 2026. So like stop doing Sora and a lot of the consumer facing stuff.

Nat Rubio-Licht: Yes, but they're not completely done with the consumer-facing stuff.

Jason Hiner: For sure.

Nat Rubio-Licht: Like, like ChatGPT Health to me feels like a consumer thing.

Jason Hiner: Yeah.

Nat Rubio-Licht: ChatGPT feels like a consumer thing.

Jason Hiner: Yeah.

Nat Rubio-Licht: My mom texted me today and she asked me for help with something on her phone because my mother is in her late sixties and sometimes doesn't know how to work her phone. And I was like, Oh, I'm sorry. I can't help you right now. Um, I just like I'm busy and that makes me sound like a terrible daughter. However, it wasn't an urgent thing. I'm so sorry. I'm so sorry, mom. Um, it makes me sound terrible. Anyways, um, but then she texted me later. She's like, Oh, I figured out how to do it. And then she says ChatGPT knows how to do it. And I'm like, Oh, but she clearly went to ChatGPT to help her with something.

Jason Hiner: Yeah.

Nat Rubio-Licht: So however, my mom is not someone that is going to start an enterprise. My mom is not going to pay 20 bucks. My mom's not even going to pay $20 a month. So what is the benefit of serving that audience?

Jason Hiner: Yeah. I mean, it's, it's the broader public. If you get more people using it, more people familiar with it, you know, they're going to need services at some point. So for, you know, broadly the bet is like, eventually say your mom needs a service to, um, look for a place to go on vacation. And there's a chat bot like her, uh, travel, you know, her favorite, you know, travel hotel or travel agent or whatever. It says like, Oh, well, we have this option. You can use this. And if she's used ChatGPT, right? Like then all of a sudden it's like, Oh, great. I used that that one time when Nat wouldn't answer my message. And I was able to, I was able to figure it out.

Nat Rubio-Licht: Right.

Jason Hiner: So it's like the, the bet is that if more people get used to using this stuff, like there are going to be a lot of uses that eventually are also businesses and we can call them enterprise in some fashion, right?

Nat Rubio-Licht: So then how does this connect back to the bubble?

Jason Hiner: So the bubble, the thing is you saw this thing on Instagram and, and then we were talking about it.

Nat Rubio-Licht: Yeah.

Jason Hiner: This is one of the things that made us want to talk about this. We're like, man, the views on the bubble are all over the place. So people were like, the bubble is dead. It was never a thing. And then then you have, you know, like you saw an Insta where it's kind of like, here it is. And this is why all of this is going to crash and burn.

Nat Rubio-Licht: Yeah. All this is going to crash and burn.

Jason Hiner: So talk a little bit about what you saw and why we were chatting about it.

Nat Rubio-Licht: I saw this video of somebody who was talking about how the economics of AI don't make any sense because they brought up that if, even though someone spends $200 on ChatGPT—sorry, not ChatGPT, but a Claude subscription, the maximum subscription you can get with Claude, it costs $8,000 a month to service that subscription.

Jason Hiner: That's what they said.

Nat Rubio-Licht: That is what they said. I don't, I, they did not cite their sources to be fair. However, the economics of that don't necessarily, the economics that this person mentioned don't necessarily add up. For one, a lot of the money is spent on, a lot of that $8,000 is spent on inference, sorry, on training rather than inference. Inference is actually cheaper than training a big powerful model.

Jason Hiner: Yeah. What they were talking about, just to, you know, interrupt you for one second, is like the fact that, you know, you can look and you can see if you pay per token, you know, when you buy that $20 subscription or the $200 subscription, you get a certain number of tokens. Like with the $200 one, you get like $8,000. In this case, they looked at one of these charts that shows like, oh, that's $8,000 worth of tokens.

Nat Rubio-Licht: Yeah.

Jason Hiner: And they're, the leap that they made that wasn't right was like, oh, it costs Anthropic $8,000 to serve those tokens, which is not the case.

Nat Rubio-Licht: Yeah.

Jason Hiner: Right. Because if you were paying per token, like enterprises do, if you pay for what's called the API, then, you know, you end up, you pay a lot more.

Nat Rubio-Licht: Yeah.

Jason Hiner: But you don't pay that when you pay the $20, you know, subscription to your consumer. Why? Because most people never get anywhere close to that.

Nat Rubio-Licht: Yeah. It's such a very, very small fraction. It's like offering an all-you-can-eat buffet and then you're not going to eat the whole buffet. Most people are going to get

Jason Hiner: a salad and one thing, right? Like, and that meal, they paid 50 bucks for, for like $7 worth of food.

Nat Rubio-Licht: Exactly.

Jason Hiner: Right. Most people, it ends up being like that. And so it was like this flawed, you know, reasoning, but you and I were interested in it because it's like, look, this is out there and people on, on Instagram are even just talking about it.

Nat Rubio-Licht: a lot of likes, a lot of shares.

Jason Hiner: Exactly. This is why it's all, this is why AI, you know, this is like normal people just talking about like why AI is a bubble and it's all going to collapse on itself.

Nat Rubio-Licht: Yeah. And I think that that is the general perception of the industry that it's all snake oil.

Jason Hiner: Yeah.

Nat Rubio-Licht: And I'm not going to say that none of it is snake oil. As you were saying today, there are plenty of companies that are going to burst and pop or plenty of companies that are model wrappers or that are all doing the same thing of like, you know, we're a control plane for agents.

Jason Hiner: Exactly. Agent harnesses. I was telling you, how many people come to me and pitch me on like, we're going to help you control your AI agents.

Nat Rubio-Licht: Exactly. And it's like, yes, every company is doing, there's a lot of companies that are doing the exact same thing, but there is also a lot of companies that are doing really cool stuff. And there is undoubtedly value here, which is why I think that we are going to land somewhere in the middle. But how do we, what does that middle look like?

Jason Hiner: Yeah, I think that this is one of the things that, that we were talking about as well as where we got to the Tinker Bell thing, which is that the bubble, as long as people believe it, right? Like that's what props it up is people believe enough people believe this is the next big thing that keeps it going, right? Where it's going to start to falter and like it's going to be natural. There's going to be a trough of disillusionment. There's the Gartner thing, right? Like the peak of heightened expectations and there's the trough of disillusionment. Like we are definitely in the height of the peak.

Nat Rubio-Licht: You don't think we've already reached—

Jason Hiner: I mean, I think we're sliding toward the trough of disillusionment 100%. Right? You look at people today in the US versus 12 months ago, their opinion has gotten worse about AI, right? So, I mean, we're sliding toward the trough of disillusionment. We're definitely not at the bottom though. But when people stop believing, when more people stop believing, that's when some of these sort of bubbles start to pop or anybody, if somebody, you know, if data comes out and it's like, oh, this company was spending $100,000 per customer and they're only charging customers, $25. Well, then, like, more videos like the one that, you know, we were talking about start to go viral, and then people are like, oh man, this whole thing is like a pyramid scheme or something, right? Like that's the stuff that that will happen. Then like every little thing that happens, right? People start getting more, more like, oh, there it is. It's a bubble. Yeah. We were sort of at that. Remember like at the end of last year, we were sort of at that. People, the bubble was like all people were talking about.

Nat Rubio-Licht: I remember.

Jason Hiner: Yeah. And then like agents happen and then there's like the hype cycle started all over again.

Nat Rubio-Licht: Yeah, it does feel like the, we've been momentarily distracted from this. But I think all the bubble conversations were largely fueled by this idea of circular financing as it related to the trillion dollars of commitments for data center deals. However, we are now coming to this point where both OpenAI and Anthropic are like desperate for compute. They don't have enough compute. They don't have enough compute.

Jason Hiner: That's the dirty little secret about Mythos was like, oh, we pulled it back because it's just too powerful. Okay. But part of it was that they just didn't have the compute to run it.

Nat Rubio-Licht: Yeah. It's pretty clear. And that's why they partnered with SpaceX and they're partnering with all of those other companies because they're like, we need to get our hands on compute right now because our, we're running the rate limit issues with our, with our clients and we're not able to serve all the people that want us.

Jason Hiner: When I say it's pretty clear, I mean just that. Like, remember they had lots of outages. Yeah. Remember after they released Mythos before the US government, you know, had them rescind it, that all of a sudden they started signing all these compute deals.

Nat Rubio-Licht: Yeah. That you mentioned.

Jason Hiner: And so it was like a little bit of a panic move, right? They didn't, they didn't have the compute to run it. And all of a sudden all the compute that OpenAI was lining up last year, all of a sudden, like, which looked insane. Now, all of a sudden, it looks a little bit more prescient, looks a little more like, okay, that makes sense. And then we also had the thing we should talk about too, is where we just came through earnings season and Google, Amazon and Microsoft released their earnings and the cloud companies are just raking it in from all of this. It's a good day to be a hyperscaler. It's an amazing time to be a hyperscaler. Yeah. They are up, you know, double digits in all cases.

Nat Rubio-Licht: I think Amazon hit a $3 trillion market cap for the first time ever. Only the fifth company ever to ever do that.

Jason Hiner: Yeah. Yeah.

Nat Rubio-Licht: So yeah, it's, and it does make me think that I think that where the buck stops is with the enterprises, because at the end of the day, yes, that is where a lot of the money is coming from. That is where a lot of the value is being derived for the model providers, for the cloud providers. That's, those are the people that are eating up everything that's being put down right now. And I do wonder if the enterprises are going to get to a point, and I think they're actually starting to approach that right now with this whole idea that we're tugging the reins on token maxing. Oh yeah. I guess the question is, are enterprises going to become disillusioned on their own? And how is that going to cause a domino effect?

Jason Hiner: Boy, yeah, we have to talk about that, because that's been a huge theme of all of the like spring, summer conference season. Yeah. It's at all these events. So you went to Microsoft Build and Snowflake Summit. I went to Databricks Data + AI Summit, Confidential Computing Summit, and others. And that's what all the enterprises were talking about. One was like control. How can we like better control this? Because they're always going to be worried about that. But then the bigger one was like cost. They're like, we're spending too much money on this, and we've got to get it under control.

Nat Rubio-Licht: Yeah.

Jason Hiner: So that means a few things for sure. That they're still not sure about the ROI, because if they were, and they were like, okay, we're using this and it's driving a lot of revenue for us, they'd be like, keep pushing the gas, which is what the hyperscalers are doing. They're like, we can't even buy enough machines to scale up to the demand we have right now, the hyperscalers. But like the enterprises, who are the ones actually buying the demand, right? They're like, throwing up the white flag. They're like, we got, we're just spending too much on AI, and we need to figure out.

Nat Rubio-Licht: I know. Yeah.

Jason Hiner: So, but then you had this question for me, because I told you about this company that I met at Ai4, as we're recording this, at the Ai4 conference. And there was this company—really, really great company. But they had told me, they shared with me, that they had not a small number of engineers who were spending 1.5 times in tokens—

Nat Rubio-Licht: Crazy.

Jason Hiner: Their total compensation.

Nat Rubio-Licht: Crazy.

Jason Hiner: So if we'll just say they were making $100,000 or $200,000—because engineers are expensive these days—we'll say they're making $200,000. That means they're spending $300,000 in tokens, annualized.

Nat Rubio-Licht: Yeah.

Jason Hiner: And you were like, what's the ROI? What are they making on it? And I was like, well, what they've told me is like, they're not focused on the unit economics. They're like, they're seeing so much acceleration, and they're still profitable. And so you're like, so what's the ROI? Like, what are the, and I'm like, I don't know.

Nat Rubio-Licht: I feel like that is also a very, that's a special case. Like, I feel like they're almost the lucky ones. Like,

Jason Hiner: exactly.

Nat Rubio-Licht: I don't think that is going to be the case for everybody. And I think actually, that's where we land somewhere in the middle. There are going to be cases like that, where there are enterprises that are like, we can spend as much money on tokens as we want, because AI is helping us ship and bring value and whatever.

Jason Hiner: That company is not public too. So that helps.

Nat Rubio-Licht: Yeah. So like, they're like, we don't care at all. We could just spend as much money. And then there are going to be other companies that are like, oh, we overbought and now the Claude bill came in, and now we are stinging a little bit and have to lay people off. Well, I think that that's where we get this, somewhere in the middle. And I think that companies are desperate to figure out that ROI equation and figure out where exactly on that spectrum they sit.

Jason Hiner: So this is where I sort of fall into the thing of like, I don't think there's one bubble. I think there are probably lots of little bubbles, right?

Nat Rubio-Licht: It's bubble wrap.

Jason Hiner: It's bubble wrap. Bubble wrap is so fun.

Nat Rubio-Licht: Yeah. Bubbles. It's sucky bubble wrap.

Jason Hiner: Is it? What? What are we, good bubble wrap?

Nat Rubio-Licht: Normal bubble wrap.

Jason Hiner: I'm with you. I'm with you. So when it comes to bubble wrap, do you like the bigger bubbles or the smaller ones?

Nat Rubio-Licht: We're getting so off track right now.

Jason Hiner: I really need to know.

Nat Rubio-Licht: I like the big bubbles.

Jason Hiner: The big bubbles.

Nat Rubio-Licht: Yeah.

Jason Hiner: So how this relates to AI is that's why people like to talk about big bubbles, right? Like, the bubble is going to pop. It's going to make, because it's more dramatic, but probably more likely it's the small bubbles, right? There are going to be lots of small bubbles where there was just like too much exuberance about that thing.

Nat Rubio-Licht: I see what you did there.

Jason Hiner: About this thing. And like those little bubbles pop and it doesn't make a whole lot of drama. And maybe, you know, it doesn't, we don't write about it in The Deep View.

Nat Rubio-Licht: So at the end of the day, we like drama.

Jason Hiner: I mean, I'm not talking about myself.

Nat Rubio-Licht: Oh, sure. Sure. At the end of the day, the industry likes drama.

Jason Hiner: I mean, it's sort of the history of storytelling altogether.

Nat Rubio-Licht: Fair enough.

Jason Hiner: People love drama. And what is, what makes a good story? Conflict, right? And, and there's got to be, you know, beginning, middle, and end and you got to resolve some conflict.

Nat Rubio-Licht: Okay. Gotcha.

Jason Hiner: Big bubbles make better stories.

Nat Rubio-Licht: Big bubbles make better stories. So it's a bunch of little bubbles.

Jason Hiner: I think it's a lot of little bubbles. Look, that doesn't mean there couldn't also still be a big bubble because some of these things, that's where the like Tinker Bell thing comes in. Because if a bunch of little bubbles start to pop and people do start to notice, then they start to lose confidence. And then people are like, oh, I need to, I need to get out of AI and I need to go back into something more stable. If, so if employees do that, if investors are like, okay, the AI thing is over, we're already seeing a little bit in the market already. You know, one of the things that investors talk about is like rotation. So like right now, people are like rotating out of chip companies, meaning they bought a lot of them. The price rose really high. Now they're like, okay, the price is just too high because now it's whatever 50 times earnings. And so it's never going to get, it's, it's just not going to be able to meet the, the, you know, the growth level for that. So they're like, okay, now I'm going to go invest in, you wrote a story on this.

Nat Rubio-Licht: Oh yeah, people investing in like the more foundational technologies, things like energy and minerals and yeah, those sorts of things, which I find to be fascinating.

Jason Hiner: That's rotation. So they rotated out of big chip companies. They rotated into energy and materials.

Nat Rubio-Licht: So wait, have they made their way down the stack then?

Jason Hiner: Yes, tell me more. Say it.

Nat Rubio-Licht: What do you mean? Have they, so I, the story that I wrote was that people have made their way, like the, the investors made their way deeper into the stack by investing in these very foundational technologies and these essentially these more stable industries. But I think the chip industry used to be that.

Jason Hiner: Yeah.

Nat Rubio-Licht: And I wonder if it went from like, oh, we're interested in like this top layer of like APIs and then models and then infrastructure and chips and now they're deeper. That's the last layer though.

Jason Hiner: Where does it go? Well, what, you know what? Does it go back? It does. It goes back. So you rotate right back to the cherry on top. This goes, you know, this one starts to go down, goes down because you rotate it into this and then that sort of gets underpriced, right? Like fewer people buy that. So then when it gets, then it gets lower and naturally like if this thing gets higher, you're like, okay, this one's probably a little too high. Now I'm going to rotate back into this. You see this in sort of, you know, stock market investing. If we, you know, there's a lot of views that stock market investing is white collar gambling, you know, and there's certainly truth to it, but you can, there's another way to look at it where it's a, it's a measure of this sentiment, right? Like of, of what people, the level of confidence people are having in certain, you know, things. And so with people, as you talked about moving into sort of lower levels of the stack, they're losing a little confidence in, in chips that they're going to continue to grow like crazy and they have more confidence that, that these other things are going to grow like crazy. And, like, what does that mean? It probably means those things, like minerals and power, are long term.

Nat Rubio-Licht: Long game.

Jason Hiner: That's a long game.

Nat Rubio-Licht: Yeah. And so not only that, those are things that, as I wrote about in that story, those are things that can benefit so far beyond AI, like investing in new kinds of sustainable energy is something, okay, yeah, that energy can be used to power data centers. That energy can also be used to power homes. So those are things that have a shelf life beyond whatever happens with this whole bubble, not bubble, deflated, popped, whatever market.

Jason Hiner: AI bubbles.

Nat Rubio-Licht: That's what I got on bubbles.

Jason Hiner: And Tinker Bell.

Nat Rubio-Licht: And Tinker Bell.

Jason Hiner: And small bubbles versus big bubbles.

Nat Rubio-Licht: Small bubbles, big bubbles, bubble wrap, Tinker Bell. And all-you-can-eat buffets.

Jason Hiner: I have a feeling we're going to probably be talking about this again.

Nat Rubio-Licht: I think so too.

Jason Hiner: All right.

Nat Rubio-Licht: All right.

Jason Hiner: Thank you.

Nat Rubio-Licht: Thank you.