#423 Jordi Visser: Why the AI Bears Are Wrong — and the One Asset He’d Bet On Right Now

Anthropic

Across the shows: 38 speakers in other episodes, 21 bullish and 41 bearish calls, past 30 days

• Jordi Visser · Visser LabsBullish

  Anthropic's revenue run-rate will reach roughly $300–400 billion.

  Visser says Anthropic's revenue run-rate has grown from about $1 billion to $10 billion and is on track toward $100 billion. He says it will not reach a trillion next year, but he thinks it will safely reach around $300 to $400 billion, driven by adoption of AI models. He estimates that together with OpenAI's $200 to $300 billion this would mean $600 to $700 billion in revenue from the two companies. He cites the progress of models and the disappearance of worries about scaling laws as support for this view.

  “They're not gonna be at 1,000,000,000,000 next year, I'm sure... But do I think they'll be somewhere around 300 to 400 safely? Yes.”at 55:17

OpenAI

Across the shows: 19 speakers in other episodes, 8 bullish and 16 bearish calls, past 30 days

• Jordi Visser · Visser LabsBullish

  OpenAI's revenue run-rate will safely reach roughly $200–300 billion.

  Visser says OpenAI will safely reach $200 to $300 billion in revenue, which together with Anthropic's expected $300 to $400 billion would mean $600 to $700 billion from the two companies. He argues this is driven by real adoption of AI models, and that continued progress in the models should keep it going. He notes that concerns about scaling laws stalling progress have already faded.

  “Do I think OpenAI will be 200 to 300 safely? Yes.”at 55:26

Qualcomm

• Jordi Visser · Visser LabsBullish

  Qualcomm should benefit as AI spreads into edge devices across homes and cars.

  Visser says semiconductors have been consolidating for six to seven months after strong gains, and he recommends moving into names like Qualcomm and AMD that benefit from edge devices. He argues AI will be installed in every product, including ovens, phones, and cars, so edge-device chipmakers should benefit as AI becomes part of everyday life.

  “you can start to move into some things like Qualcomm and AMD, things that are gonna benefit from edge devices because it's gonna be in your oven or your stove. It's gonna be in your telephone at home. It's gonna be in your car.”at 55:58

Advanced Micro Devices

Across the shows: 2 speakers in other episodes, 2 bullish and 0 bearish calls, past 30 days

• Jordi Visser · Visser LabsBullish

  AMD should benefit as AI spreads into edge devices across homes and cars.

  Visser says chips have been consolidating for six to seven months after strong gains, and he recommends moving into names like AMD and Qualcomm that benefit from edge devices. He argues AI will be installed in every product, including ovens, phones, and cars, so edge-device chipmakers should benefit as AI becomes part of everyday life.

  “you can start to move into some things like Qualcomm and AMD, things that are gonna benefit from edge devices because it's gonna be in your oven or your stove.”at 55:58

Bitcoin

Across the shows: 9 speakers in other episodes, 5 bullish and 3 bearish calls, past 30 days

• Jordi Visser · Visser LabsBullish

  Bitcoin will be the best-performing asset over the next year; crypto is cheap and under-owned.

  Visser calls Bitcoin a belief with no fundamentals, but argues that hundreds of millions of people have chosen it and that it ranks among the world's largest assets. He says it has been the best-performing asset since 2010 (or 2012) on a Sharpe basis, beaten only by the Magnificent Seven, and that its ecosystem will grow now that financial guardrails are arriving. He calls Bitcoin the S&P 500 of that world. He argues crypto is the only cheap, under-owned asset because Wall Street does not cover it, and that AI agents, which he expects to be the real users of crypto, will drive a major stage of growth. He expects it to be the best-performing asset in the next year, while noting that a 100-to-1 price on the top board makes it a favorite.

  “in my opinion, a favorite, meaning I think this will be the best performing asset in the next year, and it's on the top board at about a 100 to one right now.”at 1:05:16

AI: overall outlook

• Jordi Visser · Visser Labs

  AI progress is accelerating well beyond expectations, and it is the only macro factor he tracks now.

  AI is accelerating faster than most people expected this year, and this shows up in macro-relevant numbers. Visser says the only macro thing he cares about now is AI. Coding agents got good enough to let digital employees do the work, then personal agents that connect to apps and use a computer arrived around August and September. He argues that people who ignore AI because they are unfamiliar with it are making a mistake.

  “AI is accelerating at a pace that has gone well beyond what people expected this year, and this shows up in a variety of numbers that are really important to the macro environment.”at 2:55

• Jordi Visser · Visser Labs

  AI erodes competitive moats; every company will be disrupted as competition accelerates.

  Visser argues that all moats are disappearing. He cites Amazon's destruction of mall stocks and Netflix and Google's damage to TV and newspaper businesses as precedents. He says a moat in a business no longer protects it in a world of AI, where competition accelerates dramatically, so he expects every company to be disrupted by AI.

  “all moats are disappearing. So if you thought you had a moat in your business, it doesn't exist in a world of AI where competition accelerates dramatically.”at 36:59

• Jordi Visser · Visser Labs

  AI is the greatest innovation yet, with a major impact on health and human life.

  Visser argues that AI will have a huge impact on cancer, which affects almost everyone, and will help people avoid sudden deaths from disease. He says the path to this is already visible and that the data centers being built are necessary for it. He calls AI the greatest innovation the world has seen in terms of its impact on human life, and compares negatives to past innovations such as cars and planes.

  “AI is the greatest innovation the world has seen in terms of the impact it will have on human life.”at 59:18

AI chips & semiconductors

• Jordi Visser · Visser Labs

  Semiconductors up 60%+ YTD; consensus shorts ignore a fourth straight year of 40%+ gains.

  Semiconductors are up more than 60% year to date, and this will be the fourth year in a row they gain at least 40%. Visser says he hears people constantly talking about shorting them and cannot find anyone publicly saying they want to buy them. He reads this as a sign that bears are fighting a strong trend.

  “Semiconductors are up 60 plus percent year to date. This will be the fourth year in a row that they've been up at least 40%.”at 3:41

• Jordi Visser · Visser Labs

  Semiconductors, up 65% this year, may drift four months but should rise next year on demand.

  Visser says semiconductors are up 65% year to date and have consolidated for about six to seven months, and that some commentators think they are about to break down. He says he does not mind if they go sideways for another four months. He expects them to go up again next year because demand is there and more chips are needed.

  “they're up 65% year to date. I mean, let them go sideways for another four months. I don't really care. Like, if next year they go up again because the demand is there, we need more of them”at 58:08

US growth & recession risk

• Jordi Visser · Visser Labs

  The US economy is K-shaped: AI-linked earnings boom while rate-sensitive businesses suffer.

  Earnings are growing rapidly and margins are expanding, but businesses that depend on low rates are struggling because zero-rate refinancing let them survive longer than they should. Visser says this K-shaped split exists in the US and globally. Investors who bet on mean reversion in small caps, Europe or Asia have been hurt by it.

  “So you do have a k shaped economy, not just in The US, but I would say globally.”at 4:29

• Jordi Visser · Visser Labs

  US is in recession in rate-sensitive sectors while the rest of the economy booms, a K-shape.

  Visser says he reads the economy from the stock market, looking at the second derivative and at how sectors are performing. He argues the US is in recession in anything sensitive to interest rates, while anything not sensitive to rates is booming. Julia La Roche ties this to a K-shaped economy seen in the market's internals, and Visser agrees that breadth is weak because of this split.

  “We are in a recession in this country based on anything sensitive to interest rates. Anything not sensitive to interest rates, we are booming.”at 36:09

Consumer AI products

• Jordi Visser · Visser Labs

  Personal AI agents will soon order for users and deflate middlemen who take fees along the way.

  Visser argues that agentic commerce will drive deflationary pressure on middlemen who collect fees, and that AI native companies will use stablecoin rails and open-source models instead of traditional bank rails. He expects Siri, Alexa and Google's Spark to reach agentic ordering within about six months. Once personal agents can switch subscriptions or move savings, the disruption will happen quickly.

  “So this is the beginning of [agentic] commerce... But over the course of the next six months, I'm certain that Siri will now reach that point, that Alexa will reach that point, Google already has a version of it with Spark”at 9:13

• Jordi Visser · Visser Labs

  Muse adoption is tracking ChatGPT's early pace; ignoring AI is a costly mistake.

  Visser cites data suggesting Muse is growing as fast as ChatGPT did at launch. He says this matters whether or not people use it, because the world will change regardless. He describes his own use: the assistant read his bills and emails, helped him switch from Fios to YouTube TV, and manages his calendar and unwanted email. He sees personal assistants as early but already useful.

  “whether you use it or not, it's not important. The data already suggests that Muse is going as fast as ChatGPT did when it came out, so you're making a mistake if you don't understand how much the world is gonna change because of this.”at 9:41

US stock market outlook

• Jordi Visser · Visser Labs

  S&P 500 should be up 30%+ on near-30% earnings growth, but multiples have compressed instead.

  Earnings are growing near 30%, top-line revenue is growing 14 to 15%, and margins are expanding across the board. Visser says these conditions usually come only when the government has printed heavily, and he would have expected the S&P to be up at least 30% as a result. It has not risen that much, and multiples have compressed this year. He treats the gap between earnings and price as an unexpected development.

  “I would have expected the S and P to be up at least 30%.”at 34:49

• Jordi Visser · Visser Labs

  S&P 500 likely flat a year out, with 20–30% corrections along the way.

  Visser says that even though he is very bullish on AI, it is very likely the S&P 500 will be unchanged a year from now. He expects hiccups and 20 to 30% corrections, citing the July unwind in crowded and concentrated AI names, which he says had no clear trigger and happened only because positioning was so crowded. He notes the market has risen over time and that AI names are consolidating.

  “I believe it is very likely that a year from now, as bullish as I am on AI, that the S and P 500 is unchanged.”at 1:03:16

AI infrastructure & data centers

• Jordi Visser · Visser Labs

  Debt-funded hyperscaler AI spending will turn Magnificent Seven into low-growth, utility-like companies.

  Visser says the Magnificent Seven are spending heavily on debt to build AI infrastructure. He expects them to become utility-type companies with many data centers but little growth. He also notes that hyperscaler equity has hit all-time highs even as their debt rises, yet CDS spreads are widening because the size of the debt is massive and highly correlated, and he expects those spreads to rise further. He treats the market's credit signals as the early warning for this risk.

  “Well, now I think the mag seven are spending tons of money on debt. I think they're gonna become utility type companies that have a lot of data centers, but not a lot of growth anymore”at 38:25

• Jordi Visser · Visser Labs

  Compute demand will outstrip supply for at least three years; agentic compute is the safest investment.

  Visser argues that agentic compute is the safest place to invest for at least the next few years, and says he built an index for it because the sector cannot be looked up. The category includes semiconductors, gas turbine and battery makers such as Caterpillar and Siemens, and chemical companies. He says compute supply will not meet demand at any point in the next three years, and that demand is exponential. Digital agent adoption only began this year, and he sees the recent rise in Anthropic and OpenAI as evidence. He says the data centers being built are necessary for the path to AI's benefits.

  “we will not, at any point in the next three years, have enough supply of compute for the demand. The demand is infinite. It's exponential.”at 52:02

International equities

• Jordi Visser · Visser Labs

  Mean-reversion bets on small caps, Europe and Asia have been painful and will keep being so.

  Visser says investors who keep betting on mean reversion in small caps, Europe or Asia have been hurt, because the AI-driven earnings and multiple leadership sits in US big tech. He sees the K-shaped economy as global, which means the bet on laggard markets has not worked. He does not give a forecast for those markets.

  “Everyone that keeps trying to bet on mean reversion in small caps or in Europe or in Asia or anything along those lines, it's been painful.”at 4:34

US rates & the Fed

• Jordi Visser · Visser Labs

  The economy is no longer rate-sensitive, so the old homebuilder, bank and car playbook no longer works.

  Since 2007, corporate debt has grown less than the stock market, the Fed now has crisis facilities, and the tech-led economy has displaced the rate-sensitive sectors (homebuilders, banks, cars). Visser links this to the iPhone and the rise of tech companies, which now make up far more of the market than in 2007. He concludes that the old macro playbook built on rates no longer applies.

  “So there's no way that you cannot adapt your playbook if now the economy is just not rate sensitive the way that it was in the past when it was homebuilders and banks and cars and everything along those lines.”at 7:33

Frontier AI labs & model race

• Jordi Visser · Visser Labs

  Anthropic and OpenAI ARR could reach about $170B this year, up from $30B, with no bubble signal.

  Anthropic and OpenAI combined had about $30B of ARR at the end of last year and are expected to finish this year near $170B, a jump Visser says nobody expected. Coding agents drove the growth, and the combined revenues this year exceed the entire software stack. The two companies are worth about $3.5T combined before any IPO, up from almost nothing about fifteen months ago. Despite this, people still debate whether ARR will fall back.

  “At the end of last year, Anthropic and Open AI combined had 30,000,000,000 in ARR. They're gonna finish this year somewhere around a 170,000,000,000, so they've gone from 30,000,000,000 to a 170,000,000,000.”at 8:15

Software & SaaS

• Jordi Visser · Visser Labs

  SaaS stocks are being rerated lower as AI makes long-term cash flows uncertain, despite steady earnings.

  Software companies such as Salesforce and Adobe kept earning money while their stocks fell, because the market is valuing them on terminal value. If investors doubt that these businesses will exist in three years, the valuation has a cliff that pushes multiples down. Visser says the same compression has hit the S&P 500 this year. Salesforce's use of Anthropic has not helped its stock much.

  “these companies are getting rerated lower with multiple compression because of a concept of terminal value, which is the value of any public company is the present value of discounted cash flows going forward.”at 13:09

Banks

• Jordi Visser · Visser Labs

  Bank multiples will compress as agentic AI and crypto disrupt deposits; bank stocks below 200-day average.

  Bank stocks have declined since about September 15, and the banking index is now below its 200-day moving average, which Visser says has happened only once this year, during the Iran situation. He argues that banks earn a spread by taking deposits at zero and buying short Treasuries at about 4.5%, and that margin disappears once agents move money to higher yields or into Stripe or crypto. He expects more agentic and crypto disruption news over the next twelve months, which should compress bank multiples. He says this is more like the SaaS derating than a collapse in the banks themselves.

  “there's gonna be more and more news that's gonna come up about disruptions from the agentic side, disruptions from crypto. That's what I expect over the next twelve months. And I just think that people are going to view the multiples on the bank stocks as needing to be much lower.”at 14:53

AI bubble & tech valuations

• Jordi Visser · Visser Labs

  AI bubble warnings are wrong; historical math does not apply to a technology shift like the iPhone.

  Visser argues that bearish macro forecasts rely on historical math that does not fit a technology shift. He says the iPhone destroyed cameras, newspapers and malls and built the companies that now make up concentration in the stock market. The US is now 75% of MSCI World, and tech including Amazon, Tesla and Google is about 56% of the index versus about 13% in 2007. He says no one could have forecast Uber or Airbnb in 2008, so revenue forecasts based on current numbers are unreliable. He tracks GPU rental prices and company revenues instead, and sees no problem while those trends hold.

  “The math behind the AI thing, if you use history, is completely over the top. It's ridiculous.”at 22:29

Robotics & autonomous vehicles

• Jordi Visser · Visser Labs

  Humanoid robots will be disruptive in about three years, once AI reasoning is solved.

  Visser says the only thing holding humanoid robots back has been intelligence, specifically the ability of their brains to reason. Multimodal models and solving hard math problems have now reached that point, so embodied AI will start to affect cars and the physical economy. He says the market has only seen the coding side of AI so far, and the physical side is coming now, about three years out.

  “I say three years from now because that is around the time that humanoids will be having a disruptive effect as well, so all we're dealing with now is the coding side of things.”at 35:06

AI & jobs

• Jordi Visser · Visser Labs

  AI is hurting the job market and driving depression; US job creation is negative outside healthcare.

  Visser says that with S&P 500 earnings growing 30% this year, the US has created zero jobs, and that job numbers are negative except in healthcare. He says the corporate ladder has been destroyed, which he has seen among people who worked for him and who contact him. He argues that recent graduates with debt cannot climb out of it, and that loss of the expectation of a better future, along with housing costs, is a major source of depression. He sees AI as a big driver of this because it is affecting the job market.

  “I think people are very depressed. I think AI is a big portion of it.”at 43:58

Open-source AI models

• Jordi Visser · Visser Labs

  Open-source models cannot match frontier models, which need far more compute.

  Visser argues that the arguments that open source will replace frontier models are wrong. He says there will be two separate sides, and that the very best models, such as those from Anthropic and OpenAI, cannot be built with open source because they need more compute. He says Chinese labs have kept up with the US to some degree by distilling, or copying, US models, and that falling token prices do not apply to this top tier.

  “Cannot happen with open source. You need more compute.”at 53:22

AI regulation & safety

• Jordi Visser · Visser Labs

  Governments may try to slow AI, which is advancing faster than its safeguards.

  Visser names the ability of governments to try to slow things down as a risk to the AI bull case, citing negative side effects of the technology. He argues the technology is moving faster than the safeguards that exist, so he sees a risk of government intervention that could disrupt the thesis.

  “This technology is moving faster than the safeguards that are up.”at 56:41

Cybersecurity

• Jordi Visser · Visser Labs

  Hacking is underestimated and rising; it is the biggest risk to the AI bull case.

  Visser names hacking as the biggest risk to his AI bull case. He says people underestimate the number of hacks every day, and cites Japan declaring a state of emergency over cyber threats. He also points to a supertanker off the coast of Texas that was hacked and taken control of and driven toward a port. He believes such events happen far more often than people realize, though he does not think they change the overall AI thesis.

  “I do believe people are underestimating how many hackings there are every day.”at 56:45

Stablecoins & crypto policy

• Jordi Visser · Visser Labs

  Crypto settlement in seconds will displace slow bank wires as AI agents demand it.

  Visser argues that digital money is already here and that instant communication has outpaced money settlement, which still takes days. He says banks and the interest earned on money in transit benefit from the delay, while crypto can settle in seconds. He argues AI agents make fast settlement necessary, because agents making optimal decisions will avoid letting banks hold their money in transit. He says crypto is also safer than bank deposits, which he argues are only backed by government money printing in a crisis, and expects crypto to become a major story within a year.

  “Digital money, which is a necessity to settle things quickly, is here.”at 1:02:09