Your $20 AI Plan Costs Them Thousands. That's Not The Bubble.

AI stocks are finally getting hit, and AI stocks are finally getting hit, and you can feel the story kind of evolving you can feel the story kind of evolving you can feel the story kind of evolving in real time as they do. The tech sector in real time as they do. The tech sector in real time as they do. The tech sector is in correction territory. Big AI names is in correction territory. Big AI names is in correction territory. Big AI names are selling off. Broadcom can report are selling off. Broadcom can report are selling off. Broadcom can report record AI revenue and still get punished record AI revenue and still get punished record AI revenue and still get punished because investors wanted more. Alphabet because investors wanted more. Alphabet because investors wanted more. Alphabet and Microsoft can keep growing cloud and Microsoft can keep growing cloud and Microsoft can keep growing cloud revenue and still trade down because the revenue and still trade down because the revenue and still trade down because the market is suddenly asking the same market is suddenly asking the same market is suddenly asking the same question over and over again. Was this question over and over again. Was this question over and over again. Was this whole AI trade just a bubble? The whole AI trade just a bubble? The whole AI trade just a bubble? The spending numbers are frankly absurd. spending numbers are frankly absurd. spending numbers are frankly absurd. Google, Microsoft, Amazon, and Meta are Google, Microsoft, Amazon, and Meta are Google, Microsoft, Amazon, and Meta are all on pace to spend somewhere around all on pace to spend somewhere around all on pace to spend somewhere around $700 billion this year on AI $700 billion this year on AI $700 billion this year on AI infrastructure. Some of these companies infrastructure. Some of these companies infrastructure. Some of these companies are raising debt, some are issuing are raising debt, some are issuing are raising debt, some are issuing stock, power is tight, memory is stock, power is tight, memory is stock, power is tight, memory is expensive, data centers are taking expensive, data centers are taking expensive, data centers are taking longer to build, and inside most longer to build, and inside most longer to build, and inside most companies, the clean story that says companies, the clean story that says companies, the clean story that says this is where we get a return on all of this is where we get a return on all of this is where we get a return on all of those bucks, that's still not fully those bucks, that's still not fully those bucks, that's still not fully there. So, if you want to say this is there. So, if you want to say this is there. So, if you want to say this is starting to look like a bubble, I get starting to look like a bubble, I get starting to look like a bubble, I get it. It's not an insane reaction. But, I it. It's not an insane reaction. But, I it. It's not an insane reaction. But, I do think it is the wrong core question do think it is the wrong core question do think it is the wrong core question to ask. A stock correction will tell you to ask. A stock correction will tell you to ask. A stock correction will tell you that investors think prices are that investors think prices are that investors think prices are stretched. It doesn't automatically tell stretched. It doesn't automatically tell stretched. It doesn't automatically tell you the demand is fake. And that you the demand is fake. And that you the demand is fake. And that distinction matters a lot because the distinction matters a lot because the distinction matters a lot because the companies closest to demand are not companies closest to demand are not companies closest to demand are not pulling back. OpenAI went from roughly pulling back. OpenAI went from roughly pulling back. OpenAI went from roughly $2 billion in annualized revenue in 2023 $2 billion in annualized revenue in 2023 $2 billion in annualized revenue in 2023 to more than $20 billion and counting in to more than $20 billion and counting in to more than $20 billion and counting in 2025. Anthropic grew even faster.

2025. Anthropic grew even faster. 2025. Anthropic grew even faster. Nvidia's data center business did almost Nvidia's data center business did almost Nvidia's data center business did almost $194 billion in fiscal 2026. The $194 billion in fiscal 2026. The $194 billion in fiscal 2026. The hyperscalers are still talking about hyperscalers are still talking about hyperscalers are still talking about capacity constraints, not lack of capacity constraints, not lack of capacity constraints, not lack of demand. So, the question I think is not demand. So, the question I think is not demand. So, the question I think is not is AI a bubble, the question is which is AI a bubble, the question is which is AI a bubble, the question is which part of the AI build-out is speculative part of the AI build-out is speculative part of the AI build-out is speculative financial froth, and which part is the financial froth, and which part is the financial froth, and which part is the physical supply chain for demand that physical supply chain for demand that physical supply chain for demand that already exists. And that's what I want already exists. And that's what I want already exists. And that's what I want to separate in this video because the to separate in this video because the to separate in this video because the lazy version of the bubble argument lazy version of the bubble argument lazy version of the bubble argument compresses way too many things into one compresses way too many things into one compresses way too many things into one word. It treats inflated stock prices word. It treats inflated stock prices word. It treats inflated stock prices and aggressive private valuations and and aggressive private valuations and and aggressive private valuations and overbuilt data centers and weak overbuilt data centers and weak overbuilt data centers and weak enterprise ROI and Nvidia's revenue and enterprise ROI and Nvidia's revenue and enterprise ROI and Nvidia's revenue and OpenAI's growth and the whole future of OpenAI's growth and the whole future of OpenAI's growth and the whole future of AI as if they're all the same question, AI as if they're all the same question, AI as if they're all the same question, and they're really not. and they're really not. and they're really not. You can have a correction in AI stocks You can have a correction in AI stocks You can have a correction in AI stocks and still have a tremendous amount of and still have a tremendous amount of and still have a tremendous amount of locked-up AI demand that is not met. You locked-up AI demand that is not met. You locked-up AI demand that is not met. You can have some companies overbuild can have some companies overbuild can have some companies overbuild capacity and still have the world be capacity and still have the world be capacity and still have the world be dramatically underbuilt for inference. dramatically underbuilt for inference. dramatically underbuilt for inference. You can have weak return on investment You can have weak return on investment You can have weak return on investment in a random corporate pilot and still in a random corporate pilot and still in a random corporate pilot and still have massive demand for coding agents have massive demand for coding agents have massive demand for coding agents and research agents and customer support and research agents and customer support and research agents and customer support automation and model APIs and enterprise automation and model APIs and enterprise automation and model APIs and enterprise AI tools that actually replace hours of AI tools that actually replace hours of AI tools that actually replace hours of work.

work. work. The mistake is treating bubble as a The mistake is treating bubble as a The mistake is treating bubble as a verdict on the whole technology. It is verdict on the whole technology. It is verdict on the whole technology. It is more useful to treat the bubble concept more useful to treat the bubble concept more useful to treat the bubble concept as an invitation to map the sector as an invitation to map the sector as an invitation to map the sector because there can be bubble dynamics in because there can be bubble dynamics in because there can be bubble dynamics in the assets around AI. There can be the assets around AI. There can be the assets around AI. There can be overvaluation. There can be crowded overvaluation. There can be crowded overvaluation. There can be crowded trades. There can be data centers trades. There can be data centers trades. There can be data centers financed on assumptions that don't financed on assumptions that don't financed on assumptions that don't survive contact with reality. Some survive contact with reality. Some survive contact with reality. Some investors are going to lose money. Some investors are going to lose money. Some investors are going to lose money. Some suppliers are going to get overpaid. suppliers are going to get overpaid. suppliers are going to get overpaid. Some companies are going to build too Some companies are going to build too Some companies are going to build too much of the wrong thing in the wrong much of the wrong thing in the wrong much of the wrong thing in the wrong place. All of that can be true, but none place. All of that can be true, but none place. All of that can be true, but none of it suggests that the underlying of it suggests that the underlying of it suggests that the underlying demand is imaginary. demand is imaginary. demand is imaginary. Start with OpenAI. OpenAI has said its Start with OpenAI. OpenAI has said its Start with OpenAI. OpenAI has said its annualized revenue went from $2 billion annualized revenue went from $2 billion annualized revenue went from $2 billion in 2023 to $6 billion 2024 to more than in 2023 to $6 billion 2024 to more than in 2023 to $6 billion 2024 to more than $20 billion 2025 and growing. That's an $20 billion 2025 and growing. That's an $20 billion 2025 and growing. That's an insane growth curve, and it is the insane growth curve, and it is the insane growth curve, and it is the slowest growth curve of the slowest growth curve of the slowest growth curve of the hyperscalers. Anthropic has grown even hyperscalers. Anthropic has grown even hyperscalers. Anthropic has grown even faster from a smaller base and now is on faster from a smaller base and now is on faster from a smaller base and now is on a higher revenue run rate than OpenAI a higher revenue run rate than OpenAI a higher revenue run rate than OpenAI reportedly. And this is not about reportedly. And this is not about reportedly. And this is not about consumer curiosity. Enterprise is now consumer curiosity. Enterprise is now consumer curiosity. Enterprise is now roughly 40% of the business at OpenAI, roughly 40% of the business at OpenAI, roughly 40% of the business at OpenAI, even more at Anthropic. And companies even more at Anthropic. And companies even more at Anthropic. And companies are just lining out the door. They are just lining out the door. They are just lining out the door. They literally can't get on board it fast literally can't get on board it fast literally can't get on board it fast enough. And that matters because enough. And that matters because enough. And that matters because enterprise revenue is a lot different enterprise revenue is a lot different enterprise revenue is a lot different from I tried the chatbot once, I from I tried the chatbot once, I from I tried the chatbot once, I subscribed, and now I regret it. A subscribed, and now I regret it. A subscribed, and now I regret it. A company doesn't keep spending real company doesn't keep spending real company doesn't keep spending real budget on AI because a demo was fun. It budget on AI because a demo was fun. It budget on AI because a demo was fun. It spends because someone inside the spends because someone inside the spends because someone inside the company is passionate about this and company is passionate about this and company is passionate about this and thinks this tool is critical for code, thinks this tool is critical for code, thinks this tool is critical for code, for research, for analysis, for customer for research, for analysis, for customer for research, for analysis, for customer work, for compliance and sales and ops, work, for compliance and sales and ops, work, for compliance and sales and ops, or some workflow where the old process or some workflow where the old process or some workflow where the old process was slower or more expensive. Now, do was slower or more expensive. Now, do was slower or more expensive. Now, do some of those dollars come from some of those dollars come from some of those dollars come from companies that are chasing FOMO and

companies that are chasing FOMO and companies that are chasing FOMO and they're worried about other companies they're worried about other companies they're worried about other companies adopting AI? 100%. Does that mean that adopting AI? 100%. Does that mean that adopting AI? 100%. Does that mean that they are irrational to spend on they are irrational to spend on they are irrational to spend on intelligence inside their business? No. intelligence inside their business? No. intelligence inside their business? No. They may not use those dollars well, and They may not use those dollars well, and They may not use those dollars well, and that's why we see that forward deployed that's why we see that forward deployed that's why we see that forward deployed engineering push from both these engineering push from both these engineering push from both these companies, but the demand is there. companies, but the demand is there. companies, but the demand is there. Anthropic and OpenAI are both setting Anthropic and OpenAI are both setting Anthropic and OpenAI are both setting new records for how quickly a private new records for how quickly a private new records for how quickly a private company can grow revenue. You don't do company can grow revenue. You don't do company can grow revenue. You don't do that on a whim. that on a whim. that on a whim. It means there are paying customers in It means there are paying customers in It means there are paying customers in the system. Look at Nvidia next. the system. Look at Nvidia next. the system. Look at Nvidia next. Nvidia's fiscal 2026 data center revenue Nvidia's fiscal 2026 data center revenue Nvidia's fiscal 2026 data center revenue was about $193.7 billion. was about $193.7 billion. was about $193.7 billion. That is a very clear public signal that That is a very clear public signal that That is a very clear public signal that we have massive physical side AI demand. we have massive physical side AI demand. we have massive physical side AI demand. Those are people willing to put down Those are people willing to put down Those are people willing to put down checks for chips and systems and checks for chips and systems and checks for chips and systems and networking and memory and racks and networking and memory and racks and networking and memory and racks and commitments moving through the supply commitments moving through the supply commitments moving through the supply chain for data centers. And the chain for data centers. And the chain for data centers. And the important part is not just that Nvidia important part is not just that Nvidia important part is not just that Nvidia is selling a lot. The important part is is selling a lot. The important part is is selling a lot. The important part is what those purchases imply. Nobody buys what those purchases imply. Nobody buys what those purchases imply. Nobody buys this much AI infrastructure because this much AI infrastructure because this much AI infrastructure because they're casually experimenting with a they're casually experimenting with a they're casually experimenting with a dashboard. The chips are being bought by dashboard. The chips are being bought by dashboard. The chips are being bought by very serious boards and CEOs because very serious boards and CEOs because very serious boards and CEOs because training and inference workloads already training and inference workloads already training and inference workloads already exist, and because everyone close to the exist, and because everyone close to the exist, and because everyone close to the demand strongly believes those workloads demand strongly believes those workloads demand strongly believes those workloads are spiking fast. If you cannot have a are spiking fast. If you cannot have a are spiking fast. If you cannot have a bubble conversation at the same time as bubble conversation at the same time as bubble conversation at the same time as you have prominent leaders complaining you have prominent leaders complaining you have prominent leaders complaining about the fact that their developers are about the fact that their developers are about the fact that their developers are burning through their cloud credits too burning through their cloud credits too burning through their cloud credits too fast. Those things should not coexist, fast. Those things should not coexist, fast. Those things should not coexist, and yet this rational market, they do.

and yet this rational market, they do. and yet this rational market, they do. Now, this is where the bubble argument Now, this is where the bubble argument Now, this is where the bubble argument gets even more interesting because the gets even more interesting because the gets even more interesting because the bears are right about one thing. Revenue bears are right about one thing. Revenue bears are right about one thing. Revenue and spending don't match as neatly as and spending don't match as neatly as and spending don't match as neatly as they should yet, right? If the they should yet, right? If the they should yet, right? If the hyperscalers spend 600, 700, maybe a hyperscalers spend 600, 700, maybe a hyperscalers spend 600, 700, maybe a trillion dollars on AI infrastructure, trillion dollars on AI infrastructure, trillion dollars on AI infrastructure, you need a huge amount of future revenue you need a huge amount of future revenue you need a huge amount of future revenue to justify that investment. That's fair. to justify that investment. That's fair. to justify that investment. That's fair. You need enterprise adoption to move You need enterprise adoption to move You need enterprise adoption to move from pilots to production. You need from pilots to production. You need from pilots to production. You need agents to become reliable enough to run agents to become reliable enough to run agents to become reliable enough to run long workflows and easy enough to roll long workflows and easy enough to roll long workflows and easy enough to roll out that every company can do it. That's out that every company can do it. That's out that every company can do it. That's the critical one. You need software the critical one. You need software the critical one. You need software teams and legal teams and finance teams teams and legal teams and finance teams teams and legal teams and finance teams and support teams, everybody to change and support teams, everybody to change and support teams, everybody to change how they work. That is not a guarantee how they work. That is not a guarantee how they work. That is not a guarantee and it's going to happen in fits and and it's going to happen in fits and and it's going to happen in fits and starts. It's going to happen in an starts. It's going to happen in an starts. It's going to happen in an adoption curve. And if you're an adoption curve. And if you're an adoption curve. And if you're an investor paying a huge multiple for investor paying a huge multiple for investor paying a huge multiple for every company that touches the AI supply every company that touches the AI supply every company that touches the AI supply chain, that timing matters a lot. But chain, that timing matters a lot. But chain, that timing matters a lot. But this is exactly why the word bubble is this is exactly why the word bubble is this is exactly why the word bubble is too blunt. The question is not whether too blunt. The question is not whether too blunt. The question is not whether AI is real. The question is whether the AI is real. The question is whether the AI is real. The question is whether the cash flows arrive in the right place at cash flows arrive in the right place at cash flows arrive in the right place at the right time for the companies that the right time for the companies that the right time for the companies that are financing the buildout all the way are financing the buildout all the way are financing the buildout all the way through the supply chain. That's a much through the supply chain. That's a much through the supply chain. That's a much more interesting question. Because the more interesting question. Because the more interesting question. Because the demand can be real and the investment demand can be real and the investment demand can be real and the investment can still be poorly timed in certain can still be poorly timed in certain can still be poorly timed in certain parts of the supply chain. The parts of the supply chain. The parts of the supply chain. The technology can be absolutely technology can be absolutely technology can be absolutely transformative and some stocks can still transformative and some stocks can still transformative and some stocks can still be too pricey. The infrastructure can be be too pricey. The infrastructure can be be too pricey. The infrastructure can be necessary and some of the builders can necessary and some of the builders can necessary and some of the builders can still earn bad returns. This has all still earn bad returns. This has all still earn bad returns. This has all happened before. Railroads were real. A happened before. Railroads were real. A happened before. Railroads were real. A lot of railroad investors still got lot of railroad investors still got lot of railroad investors still got absolutely destroyed in the market even absolutely destroyed in the market even absolutely destroyed in the market even though it was a tremendous buildout and though it was a tremendous buildout and though it was a tremendous buildout and a huge success for the economy. Fiber a huge success for the economy. Fiber a huge success for the economy. Fiber was real. A lot of telecom investors was real. A lot of telecom investors was real. A lot of telecom investors still got destroyed. Cloud was real. Not still got destroyed. Cloud was real. Not still got destroyed. Cloud was real. Not every cloud-adjacent company managed to

every cloud-adjacent company managed to every cloud-adjacent company managed to capture that value. capture that value. capture that value. So when people say this looks like the So when people say this looks like the So when people say this looks like the dot-com bubble, I think the honest dot-com bubble, I think the honest dot-com bubble, I think the honest answer is maybe in some ways, but not answer is maybe in some ways, but not answer is maybe in some ways, but not the way you think it means. The dot-com the way you think it means. The dot-com the way you think it means. The dot-com bubble was notoriously decades ahead of bubble was notoriously decades ahead of bubble was notoriously decades ahead of demand. That is not what we should be demand. That is not what we should be demand. That is not what we should be seeing here. The dot-com bubble did not seeing here. The dot-com bubble did not seeing here. The dot-com bubble did not prove the internet was fake. It proved prove the internet was fake. It proved prove the internet was fake. It proved that markets can overprice the that markets can overprice the that markets can overprice the first-order version of a real platform first-order version of a real platform first-order version of a real platform shift. It's not that nothing is shift. It's not that nothing is shift. It's not that nothing is happening. The risk is that the market happening. The risk is that the market happening. The risk is that the market prices every AI-exposed asset as if it's prices every AI-exposed asset as if it's prices every AI-exposed asset as if it's automatically going to magically capture automatically going to magically capture automatically going to magically capture value. It It won't, right? Some value. It It won't, right? Some value. It It won't, right? Some companies will provide commoditized companies will provide commoditized companies will provide commoditized inputs, some are going to get squeezed, inputs, some are going to get squeezed, inputs, some are going to get squeezed, some are going to build way too far some are going to build way too far some are going to build way too far ahead of demand, some will have the ahead of demand, some will have the ahead of demand, some will have the right thesis and the wrong balance right thesis and the wrong balance right thesis and the wrong balance sheet. But the buildout itself, that's sheet. But the buildout itself, that's sheet. But the buildout itself, that's not a hallucination. The reason is not a hallucination. The reason is not a hallucination. The reason is inference. And this is the part of the inference. And this is the part of the inference. And this is the part of the AI spending story that feels very under AI spending story that feels very under AI spending story that feels very under explained on Main Street and Wall Street explained on Main Street and Wall Street explained on Main Street and Wall Street to me. Training a model is expensive, to me. Training a model is expensive, to me. Training a model is expensive, but it's episodic. You build a huge but it's episodic. You build a huge but it's episodic. You build a huge cluster, you run a training job, you cluster, you run a training job, you cluster, you run a training job, you produce a model, and then you move on to produce a model, and then you move on to produce a model, and then you move on to the next generation. Inference is the next generation. Inference is the next generation. Inference is different. Inference is the model different. Inference is the model different. Inference is the model running every time someone uses it.

running every time someone uses it. running every time someone uses it. Every prompt, every agent step, every Every prompt, every agent step, every Every prompt, every agent step, every tool call, every retry, every long tool call, every retry, every long tool call, every retry, every long context window, every document, every context window, every document, every context window, every document, every code base, every verification pass. When code base, every verification pass. When code base, every verification pass. When AI was mostly chat, inference looked AI was mostly chat, inference looked AI was mostly chat, inference looked super manageable. And that was not that super manageable. And that was not that super manageable. And that was not that long ago. That was like 7-8 months ago. long ago. That was like 7-8 months ago. long ago. That was like 7-8 months ago. A person asks a question, and waits, and A person asks a question, and waits, and A person asks a question, and waits, and reads, and maybe asks another one. And a reads, and maybe asks another one. And a reads, and maybe asks another one. And a lot of the build-out was around training lot of the build-out was around training lot of the build-out was around training runs. Agents in the last 6 months runs. Agents in the last 6 months runs. Agents in the last 6 months changed that math fundamentally forever. changed that math fundamentally forever. changed that math fundamentally forever. An agent does not ask one question and An agent does not ask one question and An agent does not ask one question and stop. It goes, right? It loops, it reads stop. It goes, right? It loops, it reads stop. It goes, right? It loops, it reads files, it calls tools, it writes code, files, it calls tools, it writes code, files, it calls tools, it writes code, it checks the result, it fixes the it checks the result, it fixes the it checks the result, it fixes the failure, it asks another model to review failure, it asks another model to review failure, it asks another model to review the input, it searches again, it runs the input, it searches again, it runs the input, it searches again, it runs again, it burns tokens over and over. again, it burns tokens over and over. again, it burns tokens over and over. That's not a conversation. That's a That's not a conversation. That's a That's not a conversation. That's a production job that runs into the production job that runs into the production job that runs into the millions and billions of tokens really, millions and billions of tokens really, millions and billions of tokens really, really fast. And once you see AI work really fast. And once you see AI work really fast. And once you see AI work that way, you can't see it any other that way, you can't see it any other that way, you can't see it any other way, and the infrastructure build-out way, and the infrastructure build-out way, and the infrastructure build-out starts to make a ton more sense. Because starts to make a ton more sense. Because starts to make a ton more sense. Because any given agent run can be thousands of any given agent run can be thousands of any given agent run can be thousands of times the inference cost of a chat times the inference cost of a chat times the inference cost of a chat conversation. Tokens are not magic, conversation. Tokens are not magic, conversation. Tokens are not magic, they're manufactured. Behind every they're manufactured. Behind every they're manufactured. Behind every answer, behind every agent tool call, is answer, behind every agent tool call, is answer, behind every agent tool call, is a physical production system. Chips and a physical production system. Chips and a physical production system. Chips and memory and networking and power and memory and networking and power and memory and networking and power and cooling and land and construction and cooling and land and construction and cooling and land and construction and ops.

ops. ops. And that's why the CapEx numbers are And that's why the CapEx numbers are And that's why the CapEx numbers are getting very serious. AI makes the most getting very serious. AI makes the most getting very serious. AI makes the most valuable software companies in the world valuable software companies in the world valuable software companies in the world look industrial today. Microsoft and look industrial today. Microsoft and look industrial today. Microsoft and Google and Amazon and Meta don't just Google and Amazon and Meta don't just Google and Amazon and Meta don't just ship features anymore. They're building ship features anymore. They're building ship features anymore. They're building factories for inference, and factories factories for inference, and factories factories for inference, and factories are expensive. They require upfront are expensive. They require upfront are expensive. They require upfront capital and utilization and supply capital and utilization and supply capital and utilization and supply assurance. They require power contracts, assurance. They require power contracts, assurance. They require power contracts, they require depreciation schedules and they require depreciation schedules and they require depreciation schedules and routing and batching and caching and routing and batching and caching and routing and batching and caching and efficiency improvements, so expensive efficiency improvements, so expensive efficiency improvements, so expensive compute is not wasted on cheap work. compute is not wasted on cheap work. compute is not wasted on cheap work. That is the real operating question. That is the real operating question. That is the real operating question. It's not is AI a bubble? The operating It's not is AI a bubble? The operating It's not is AI a bubble? The operating question is this: Are expensive tokens question is this: Are expensive tokens question is this: Are expensive tokens being spent on work valuable enough to being spent on work valuable enough to being spent on work valuable enough to justify them? That's the question of justify them? That's the question of justify them? That's the question of 2026. And that question separates what's 2026. And that question separates what's 2026. And that question separates what's real in this AI explosion of demand from real in this AI explosion of demand from real in this AI explosion of demand from what's fake very quickly. A coding agent what's fake very quickly. A coding agent what's fake very quickly. A coding agent that saves an engineering team days of that saves an engineering team days of that saves an engineering team days of work can justify that expensive work can justify that expensive work can justify that expensive inference. And these days they're saving inference. And these days they're saving inference. And these days they're saving weeks and months sometimes. A legal weeks and months sometimes. A legal weeks and months sometimes. A legal review agent that processes thousands of review agent that processes thousands of review agent that processes thousands of contracts can justify expensive contracts can justify expensive contracts can justify expensive inference. A customer service system inference. A customer service system inference. A customer service system that resolves real tickets and reduces that resolves real tickets and reduces that resolves real tickets and reduces escalation, you can justify expensive escalation, you can justify expensive escalation, you can justify expensive inference that way.

inference that way. inference that way. Now, a random enterprise chatbot on the Now, a random enterprise chatbot on the Now, a random enterprise chatbot on the website that answers shallow questions website that answers shallow questions website that answers shallow questions from a stale knowledge base and provides from a stale knowledge base and provides from a stale knowledge base and provides a terrible customer experience, not a terrible customer experience, not a terrible customer experience, not really justifying your inference there. really justifying your inference there. really justifying your inference there. And that's why the enterprise ROI data And that's why the enterprise ROI data And that's why the enterprise ROI data looks like a complete mess right now. AI looks like a complete mess right now. AI looks like a complete mess right now. AI is not one single thing. It's a is not one single thing. It's a is not one single thing. It's a general-purpose technology. It's a general-purpose technology. It's a general-purpose technology. It's a thousand different workflows with thousand different workflows with thousand different workflows with different economics. Some are really different economics. Some are really different economics. Some are really useful. Some are terrible ideas. Some useful. Some are terrible ideas. Some useful. Some are terrible ideas. Some save time at the individual level but save time at the individual level but save time at the individual level but never make it into the P&L. Some will never make it into the P&L. Some will never make it into the P&L. Some will look impressive in a demo and collapse look impressive in a demo and collapse look impressive in a demo and collapse when you put them inside a real workflow when you put them inside a real workflow when you put them inside a real workflow with permissions and exceptions and with permissions and exceptions and with permissions and exceptions and messy data and accountability. messy data and accountability. messy data and accountability. None of this is proof of a bubble. This None of this is proof of a bubble. This None of this is proof of a bubble. This is proof that adoption is uneven. And is proof that adoption is uneven. And is proof that adoption is uneven. And frankly, the companies don't necessarily frankly, the companies don't necessarily frankly, the companies don't necessarily know what to do with the new know what to do with the new know what to do with the new general-purpose technology yet. general-purpose technology yet. general-purpose technology yet. And frankly, it should be uneven. Most And frankly, it should be uneven. Most And frankly, it should be uneven. Most companies are bad at process change. companies are bad at process change. companies are bad at process change. They were bad at software implementation They were bad at software implementation They were bad at software implementation before AI. They were bad at data before AI. They were bad at data before AI. They were bad at data projects before AI. They were bad at projects before AI. They were bad at projects before AI. They were bad at cloud migration before AI. And now, cloud migration before AI. And now, cloud migration before AI. And now, they're bad at AI transformation. And they're bad at AI transformation. And they're bad at AI transformation. And we're all acting surprised. The we're all acting surprised. The we're all acting surprised. The technology can be real while companies technology can be real while companies technology can be real while companies struggle with change management. And struggle with change management. And struggle with change management. And that's where I think the better mental that's where I think the better mental that's where I think the better mental model is. It's not bubble versus no model is. It's not bubble versus no model is. It's not bubble versus no bubble. It's build-out versus payback.

bubble. It's build-out versus payback. bubble. It's build-out versus payback. The build-out is real. The demand The build-out is real. The demand The build-out is real. The demand signals are real, the constraints are signals are real, the constraints are signals are real, the constraints are also real. OpenAI's revenue is real, so also real. OpenAI's revenue is real, so also real. OpenAI's revenue is real, so is Anthropic's. Nvidia's data center is Anthropic's. Nvidia's data center is Anthropic's. Nvidia's data center revenue is real, hyperscaler capex is revenue is real, hyperscaler capex is revenue is real, hyperscaler capex is also real, and capacity constraints are also real, and capacity constraints are also real, and capacity constraints are real. real. real. The payback is the open question The payback is the open question The payback is the open question circling around all of those facts. circling around all of those facts. circling around all of those facts. Who gets paid back and when? How fast do Who gets paid back and when? How fast do Who gets paid back and when? How fast do they get paid? At what margin? On which they get paid? At what margin? On which they get paid? At what margin? On which workloads does it matter that they get workloads does it matter that they get workloads does it matter that they get paid back? How much pricing power do the paid back? How much pricing power do the paid back? How much pricing power do the hyperscalers have when they are setting hyperscalers have when they are setting hyperscalers have when they are setting prices for tokens and for workflows? prices for tokens and for workflows? prices for tokens and for workflows? This is where the market ought to be This is where the market ought to be This is where the market ought to be more thoughtful, frankly. If you're more thoughtful, frankly. If you're more thoughtful, frankly. If you're buying every AI stock because AI is the buying every AI stock because AI is the buying every AI stock because AI is the future, you're not really doing due future, you're not really doing due future, you're not really doing due diligence and analysis. You're buying a diligence and analysis. You're buying a diligence and analysis. You're buying a narrative, and narratives can flip on a narrative, and narratives can flip on a narrative, and narratives can flip on a dime. dime. dime. But if you're dismissing the entire But if you're dismissing the entire But if you're dismissing the entire thing because stocks corrected, you are thing because stocks corrected, you are thing because stocks corrected, you are also not doing analysis. You are also not doing analysis. You are also not doing analysis. You are reacting to price action and pretending reacting to price action and pretending reacting to price action and pretending it is insight. it is insight. it is insight. The useful middle ground requires a lot The useful middle ground requires a lot The useful middle ground requires a lot more due diligence and thoughtfulness, more due diligence and thoughtfulness, more due diligence and thoughtfulness, and it's much harder and rarer. It says and it's much harder and rarer. It says and it's much harder and rarer. It says AI is a real platform shift that is so AI is a real platform shift that is so AI is a real platform shift that is so transformative that there are a bunch of transformative that there are a bunch of transformative that there are a bunch of local bubble dynamics frothing around local bubble dynamics frothing around local bubble dynamics frothing around it. That means prices can fall and it. That means prices can fall and it. That means prices can fall and technology can keep advancing. It means technology can keep advancing. It means technology can keep advancing. It means some infrastructure might be overbuilt some infrastructure might be overbuilt some infrastructure might be overbuilt while other kinds of capacity remain while other kinds of capacity remain while other kinds of capacity remain profoundly scarce. It means some profoundly scarce. It means some profoundly scarce. It means some companies will spend too much and still companies will spend too much and still companies will spend too much and still not spend enough in the exact place that not spend enough in the exact place that not spend enough in the exact place that matters. Yes, that can be true. It means matters. Yes, that can be true. It means matters. Yes, that can be true. It means the biggest winners may not be the the biggest winners may not be the the biggest winners may not be the companies with the loudest AI story companies with the loudest AI story companies with the loudest AI story today. They may be the ones that control today. They may be the ones that control today. They may be the ones that control the bottleneck or own the customer

the bottleneck or own the customer the bottleneck or own the customer workflow or route inference more workflow or route inference more workflow or route inference more efficiently or turn agent output into efficiently or turn agent output into efficiently or turn agent output into durable business value. durable business value. durable business value. This is the distinction I would watch. This is the distinction I would watch. This is the distinction I would watch. Don't ask if a company is doing AI if Don't ask if a company is doing AI if Don't ask if a company is doing AI if you're trying to figure out investments you're trying to figure out investments you're trying to figure out investments here. And none of this is investment here. And none of this is investment here. And none of this is investment advice. Ask where the demand is showing advice. Ask where the demand is showing advice. Ask where the demand is showing up. Is it paid usage or is it just up. Is it paid usage or is it just up. Is it paid usage or is it just engagement? Is it production workloads engagement? Is it production workloads engagement? Is it production workloads or is it just a pilot that got dressed or is it just a pilot that got dressed or is it just a pilot that got dressed up in a press release? Is it improving a up in a press release? Is it improving a up in a press release? Is it improving a workflow with really clear economics or workflow with really clear economics or workflow with really clear economics or is it creating more work for humans to is it creating more work for humans to is it creating more work for humans to review? Is the company buying capacity review? Is the company buying capacity review? Is the company buying capacity because customers are waiting or because because customers are waiting or because because customers are waiting or because the board wants an AI strategy? Is the the board wants an AI strategy? Is the the board wants an AI strategy? Is the model being used where expensive model being used where expensive model being used where expensive reasoning matters or is it just premium reasoning matters or is it just premium reasoning matters or is it just premium compute being burned on cheap tasks? compute being burned on cheap tasks? compute being burned on cheap tasks? Those questions are a lot less dramatic Those questions are a lot less dramatic Those questions are a lot less dramatic than bubble or revolution, but they're a than bubble or revolution, but they're a than bubble or revolution, but they're a lot more useful for determining where lot more useful for determining where lot more useful for determining where investment dollars ought to go. investment dollars ought to go. investment dollars ought to go. And they also explain why the stock And they also explain why the stock And they also explain why the stock correction does not remotely settle the correction does not remotely settle the correction does not remotely settle the issue. Markets can correct and then issue. Markets can correct and then issue. Markets can correct and then uncorrect for lots of reasons, right? uncorrect for lots of reasons, right? uncorrect for lots of reasons, right? Valuation stretch, trades get crowded, Valuation stretch, trades get crowded, Valuation stretch, trades get crowded, expectations get too high, capital expectations get too high, capital expectations get too high, capital rotates, rotates, rotates, financing costs matter. A company can financing costs matter. A company can financing costs matter. A company can disappoint investors even while the disappoint investors even while the disappoint investors even while the underlying business grows, and that is underlying business grows, and that is underlying business grows, and that is what makes this moment so profoundly what makes this moment so profoundly what makes this moment so profoundly tricky. It's not that the correction is tricky. It's not that the correction is tricky. It's not that the correction is meaningless. It's telling you that meaningless. It's telling you that meaningless. It's telling you that investors are having some feelings in investors are having some feelings in investors are having some feelings in the middle of a global energy crisis the middle of a global energy crisis the middle of a global energy crisis about underwriting unlimited AI spending about underwriting unlimited AI spending about underwriting unlimited AI spending without asking harder questions. That's without asking harder questions. That's without asking harder questions. That's great. They should ask harder questions.

great. They should ask harder questions. great. They should ask harder questions. But the correction is not proof that the But the correction is not proof that the But the correction is not proof that the buildout is fake. It's proof that, you buildout is fake. It's proof that, you buildout is fake. It's proof that, you know, maybe the easy phase of this trade know, maybe the easy phase of this trade know, maybe the easy phase of this trade is over. And the next phase is going to is over. And the next phase is going to is over. And the next phase is going to require a little bit more discrimination require a little bit more discrimination require a little bit more discrimination from people who are trying to figure out from people who are trying to figure out from people who are trying to figure out investment dollars. The market will investment dollars. The market will investment dollars. The market will start separating companies with real AI start separating companies with real AI start separating companies with real AI revenue from companies with AI language revenue from companies with AI language revenue from companies with AI language in the deck. It's going to separate in the deck. It's going to separate in the deck. It's going to separate infrastructure bottlenecks from infrastructure bottlenecks from infrastructure bottlenecks from commodity exposure. It'll separate tools commodity exposure. It'll separate tools commodity exposure. It'll separate tools that create measurable workflow value that create measurable workflow value that create measurable workflow value from tools that create demo value. It'll from tools that create demo value. It'll from tools that create demo value. It'll separate companies that can finance the separate companies that can finance the separate companies that can finance the buildout from companies that need the buildout from companies that need the buildout from companies that need the buildout to be financed by someone else. buildout to be financed by someone else. buildout to be financed by someone else. And yes, it will separate SaaS companies And yes, it will separate SaaS companies And yes, it will separate SaaS companies that have sticky services that are still that have sticky services that are still that have sticky services that are still valuable in the era of agents from ones valuable in the era of agents from ones valuable in the era of agents from ones that don't. And that's super healthy, that don't. And that's super healthy, that don't. And that's super healthy, and it's exactly what should happen in a and it's exactly what should happen in a and it's exactly what should happen in a real platform shift. The first phase is real platform shift. The first phase is real platform shift. The first phase is almost always narrative. Everyone piles almost always narrative. Everyone piles almost always narrative. Everyone piles into the obvious names. The second phase into the obvious names. The second phase into the obvious names. The second phase is correction. The market realizes the is correction. The market realizes the is correction. The market realizes the story is more expensive, it's slower, story is more expensive, it's slower, story is more expensive, it's slower, it's messier than the headlines may be it's messier than the headlines may be it's messier than the headlines may be applied. So, when someone asks, is AI a applied. So, when someone asks, is AI a applied. So, when someone asks, is AI a bubble? My answer is parts of it are, bubble? My answer is parts of it are, bubble? My answer is parts of it are, yeah. If you can release a press release yeah. If you can release a press release yeah. If you can release a press release and get a 500% pop in your stock, which and get a 500% pop in your stock, which and get a 500% pop in your stock, which I've seen once or twice, those are I've seen once or twice, those are I've seen once or twice, those are examples of a bubble. But, that doesn't examples of a bubble. But, that doesn't examples of a bubble. But, that doesn't mean the whole system is a bubble at mean the whole system is a bubble at mean the whole system is a bubble at all, remotely. Sure, some valuations are all, remotely. Sure, some valuations are all, remotely. Sure, some valuations are stretched, some spending will be wasted, stretched, some spending will be wasted, stretched, some spending will be wasted, some private market marks are probably some private market marks are probably some private market marks are probably ridiculous, some companies are ridiculous, some companies are ridiculous, some companies are pretending a thin wrapper is a business, pretending a thin wrapper is a business, pretending a thin wrapper is a business, the seed rounds are getting really the seed rounds are getting really the seed rounds are getting really pricey, the valley some enterprises are pricey, the valley some enterprises are pricey, the valley some enterprises are buying expensive tools without changing buying expensive tools without changing buying expensive tools without changing the workflow enough to get the value, the workflow enough to get the value, the workflow enough to get the value, but the broader build-out is not hype

but the broader build-out is not hype but the broader build-out is not hype floating above reality. There's real floating above reality. There's real floating above reality. There's real demand underneath. There's real revenue demand underneath. There's real revenue demand underneath. There's real revenue underneath. There's real physical underneath. There's real physical underneath. There's real physical scarcity underneath. And so, the better scarcity underneath. And so, the better scarcity underneath. And so, the better question it's not when the bubble pops. question it's not when the bubble pops. question it's not when the bubble pops. I don't think that's coming. The better I don't think that's coming. The better I don't think that's coming. The better question is who survives the sorting question is who survives the sorting question is who survives the sorting that is going to happen. Because if that is going to happen. Because if that is going to happen. Because if intelligence becomes a production intelligence becomes a production intelligence becomes a production system, then the winners are not just system, then the winners are not just system, then the winners are not just the companies with the best demos. the companies with the best demos. the companies with the best demos. They're the companies that can turn They're the companies that can turn They're the companies that can turn demand into reliable, affordable, high demand into reliable, affordable, high demand into reliable, affordable, high utilization inference. They can route utilization inference. They can route utilization inference. They can route the right task to the right model. They the right task to the right model. They the right task to the right model. They can get power. They can get memory. They can get power. They can get memory. They can get power. They can get memory. They can build capacity. They can make agents can build capacity. They can make agents can build capacity. They can make agents useful enough that customers keep paying useful enough that customers keep paying useful enough that customers keep paying after the novelty wears off. That's a after the novelty wears off. That's a after the novelty wears off. That's a much harder game than the stock chart much harder game than the stock chart much harder game than the stock chart made it look last year. But, it's also a made it look last year. But, it's also a made it look last year. But, it's also a much more serious game. much more serious game. much more serious game. A bubble is hype detached from reality. A bubble is hype detached from reality. A bubble is hype detached from reality. That's kind of the definition of a That's kind of the definition of a That's kind of the definition of a bubble, right? The famous South Sea bubble, right? The famous South Sea bubble, right? The famous South Sea bubble was all about the hype. It cost bubble was all about the hype. It cost bubble was all about the hype. It cost Isaac Newton his fortune. This is Isaac Newton his fortune. This is Isaac Newton his fortune. This is messier than that. This is a real messier than that. This is a real messier than that. This is a real build-out with speculative money piled build-out with speculative money piled build-out with speculative money piled over the top, and the correction is the over the top, and the correction is the over the top, and the correction is the market starting to ask which layer is market starting to ask which layer is market starting to ask which layer is which, and that's fantastic. And that's which, and that's fantastic. And that's which, and that's fantastic. And that's a good question to keep in front of you.

a good question to keep in front of you. a good question to keep in front of you. You should be asking, where is the paid You should be asking, where is the paid You should be asking, where is the paid demand? Where's the bottleneck really? demand? Where's the bottleneck really? demand? Where's the bottleneck really? Who captures value when the tokens get Who captures value when the tokens get Who captures value when the tokens get cheaper? Is this business able to cheaper? Is this business able to cheaper? Is this business able to finance itself? Uh is the work really finance itself? Uh is the work really finance itself? Uh is the work really moving to agents in association with moving to agents in association with moving to agents in association with this particular company? And that's this particular company? And that's this particular company? And that's where the real story is. So, if you are where the real story is. So, if you are where the real story is. So, if you are worried about a particular movement in worried about a particular movement in worried about a particular movement in the stocks, I want you to keep these the stocks, I want you to keep these the stocks, I want you to keep these questions in mind. These are questions questions in mind. These are questions questions in mind. These are questions that are evergreen questions. You can that are evergreen questions. You can that are evergreen questions. You can come back to them next month and the come back to them next month and the come back to them next month and the month after. We will take a while as a month after. We will take a while as a month after. We will take a while as a market to go through this sorting. market to go through this sorting. market to go through this sorting. Remember, AI is a marathon, it's not a Remember, AI is a marathon, it's not a Remember, AI is a marathon, it's not a sprint. The business of putting AI into sprint. The business of putting AI into sprint. The business of putting AI into companies and installing it is a 10-20 companies and installing it is a 10-20 companies and installing it is a 10-20 year exercise. We're just at the year exercise. We're just at the year exercise. We're just at the beginning of that. We are writing the beginning of that. We are writing the beginning of that. We are writing the first chapter. Think about that the next first chapter. Think about that the next first chapter. Think about that the next time you look at your Robinhood account time you look at your Robinhood account time you look at your Robinhood account or think about that the next time you or think about that the next time you or think about that the next time you think about where stocks are at. That think about where stocks are at. That think about where stocks are at. That provides a larger picture and I think provides a larger picture and I think provides a larger picture and I think it's a healthier picture because AI is it's a healthier picture because AI is it's a healthier picture because AI is here for the long term. AI is the most here for the long term. AI is the most here for the long term. AI is the most transformative technology of our lives transformative technology of our lives transformative technology of our lives and it can still have a ton of froth and it can still have a ton of froth and it can still have a ton of froth around the edges while that remains around the edges while that remains around the edges while that remains true. I hope this has been helpful. You true. I hope this has been helpful. You true. I hope this has been helpful. You can follow me for more relatively sober can follow me for more relatively sober can follow me for more relatively sober takes in a world that likes to argue takes in a world that likes to argue takes in a world that likes to argue about big, big overcorrections and about big, big overcorrections and about big, big overcorrections and binaries. I do not believe the world is binaries. I do not believe the world is binaries. I do not believe the world is a light switch world, right? I don't a light switch world, right? I don't a light switch world, right? I don't believe it's either bubble or not and believe it's either bubble or not and believe it's either bubble or not and I'm going to argue against that a lot I'm going to argue against that a lot I'm going to argue against that a lot because I think it's a lazy question because I think it's a lazy question because I think it's a lazy question that underscores how little people that underscores how little people that underscores how little people understand how powerful AI actually is understand how powerful AI actually is understand how powerful AI actually is as well as how little people understand

as well as how little people understand as well as how little people understand the complexities the complexities the complexities of the AI dynamic. Yes, there are of the AI dynamic. Yes, there are of the AI dynamic. Yes, there are absolutely places in a build out this absolutely places in a build out this absolutely places in a build out this big where capital is wasted and yes, big where capital is wasted and yes, big where capital is wasted and yes, that does not mean that we don't see that does not mean that we don't see that does not mean that we don't see absolutely massive unmet demand with AI. absolutely massive unmet demand with AI. absolutely massive unmet demand with AI. Both can be true once. Demand more of Both can be true once. Demand more of Both can be true once. Demand more of your investors. Demand more of your your investors. Demand more of your your investors. Demand more of your analysis. Demand more of the markets in analysis. Demand more of the markets in analysis. Demand more of the markets in understanding how these companies understanding how these companies understanding how these companies actually work. And frankly, if you're in actually work. And frankly, if you're in actually work. And frankly, if you're in financial press, feel free to get in financial press, feel free to get in financial press, feel free to get in touch with me because I feel like this touch with me because I feel like this touch with me because I feel like this is often incorrectly reported, is often incorrectly reported, is often incorrectly reported, especially that inference piece. I don't especially that inference piece. I don't especially that inference piece. I don't think people properly understand that think people properly understand that think people properly understand that one. All right, I will see you in the one. All right, I will see you in the one. All right, I will see you in the comments. Let me know what you think. comments. Let me know what you think. comments. Let me know what you think. Sound off on where you think some of Sound off on where you think some of Sound off on where you think some of these companies are at. I would love to these companies are at. I would love to these companies are at. I would love to hear because we can argue about that. I hear because we can argue about that. I hear because we can argue about that. I think that's a healthy conversation to think that's a healthy conversation to think that's a healthy conversation to have and that would be a way for us to have and that would be a way for us to have and that would be a way for us to talk as a community about which talk as a community about which talk as a community about which companies are sorting where and why. companies are sorting where and why. companies are sorting where and why. All right. I'll see you next time. All right. I'll see you next time. All right. I'll see you next time. Cheers.