The Intangible Economy with Kai Wu

Azeem Azhar joins Kai Wu to break down the real economics of the AI boom, including the $110 billion demand base, where profits may accrue across chips, hosting, foundation models and applications, and whether spending can translate into enterprise productivity. They discuss AI infrastructure bottlenecks, open-source competition, vertical integration, organizational redesign, software moats, human judgment and the signals investors can use to identify companies turning AI adoption into durable competitive advantage.The State of the AI Economyhttps://intelligence.exponentialview.co/assets/ev-state-of-ai-economy-2026.pdfWhy AI Isn't Showing Up on Your Bottom Linehttps://www.exponentialview.co/p/why-ai-isnt-showing-up-on-your-bottom-lineAzeem Azhar on Xhttps://x.com/azeemExponential Viewhttps://www.exponentialview.co/Topics CoveredThe size and growth rate of real generative AI demandHow the AI stack divides between chips, hosting, foundation models and applicationsWhy memory and energized data centers may be the key AI infrastructure bottlenecksOpen-source models, proprietary pricing and enterprise assuranceVertical integration and foundation model labs moving into applicationsHow AI value could flow to consumers rather than infrastructure providersWhy AI productivity requires workflow and organizational redesignWhat investors can learn from earnings calls, hiring and enterprise spendingForward-deployed engineers, consulting firms and vendor lock-inWhich intangible business moats strengthen or weaken as intelligence becomes abundantTimestamps00:00 The economics and sustainability of the AI boom06:34 Mapping the four layers of the AI stack10:43 Vertical integration and cross-stack competition15:31 Why memory is becoming an AI infrastructure bottleneck20:01 Open-source models versus proprietary AI24:36 Why foundation model labs are moving up and down the stack28:51 Could AI profits become consumer surplus?33:00 Why more copilots cannot create an AI-native company37:17 Job postings and the intangible investments behind AI adoption44:16 Can forward-deployed engineers transform legacy companies?49:15 Which business moats strengthen or weaken in the AI economy?54:20 Do foundation models really have network effects?59:00 Why judgment, verification and human provenance become more valuable01:04:56 The exponential gap in data centers and education01:10:06 How Azeem uses AI to deepen research and generate ideasLearn more about the Excess Returns podcast network:https://excessreturns.coNo information discussed in this podcast should be construed as investment advice. Securities discussed may be held by the hosts and guests, their firms or their clients.

Professor Aswath Damodaran joins The Intangible Economy to break down how to value SpaceX, AI companies, intangible assets, and the future of value investing.We discuss why big markets do not automatically create big value, how AI CapEx is changing the character of major technology companies, and why the best investment stories still have to connect to the numbers.Aswath Damodaran on Xhttps://x.com/AswathDamodaranMusings on Marketshttps://aswathdamodaran.blogspot.com/Revisiting the SpaceX Valuation: A Post-Prospectus Updatehttps://aswathdamodaran.blogspot.com/2026/06/a-weeks-ago-i-assessed-value-of-spacex.htmlThe Big Market Delusion: Valuation and Investment Implicationshttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=3501688Valuing Cyclical and Commodity Companieshttps://people.stern.nyu.edu/adamodar/pdfiles/papers/commodity.pdfValue Investing: Requiem, Rebirth or Reincarnation?https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3779481Topics covered:Valuing SpaceX after its IPO and why price matters even for great companiesHow Starlink, space launch, and xAI fit into SpaceX’s valuation storyWhy total addressable market can mislead investors in AI and other disruptive industriesThe problem with AI unit economics, data centers, power, water, and reinvestment needsWhy growth can destroy value when margins and returns on capital are weakHow intangible assets, R&D, future growth, and narratives should show up in valuationThe Big Market Delusion and how overconfidence drives boom and bust cyclesWhy AI CapEx is different from the dot-com boom and could create broader risksHow AI is changing the character of the Magnificent Seven and semiconductor companiesWhy value investing became rigid, ritualistic, and righteous, and how it can evolveTimestamps:00:00 Why great companies can still be bad investments01:03 Introducing Aswath Damodaran and The Intangible Economy01:49 SpaceX IPO, Starlink, xAI, and the challenge of valuing uncertainty05:31 Why Starlink became the core of SpaceX’s current revenue10:31 How Damodaran valued SpaceX across launch, connectivity, and AI14:07 Why AI’s huge market may still have difficult unit economics17:10 The tension between SpaceX competing in AI and renting data centers to competitors20:00 Why valuation should use distributions instead of false precision22:39 How stories and numbers work together in valuation26:45 Why investors confuse promises, potential, and businesses30:49 The Big Market Delusion and overconfidence in AI investing33:02 Why the AI CapEx boom is different from the dot-com bubble35:17 How AI infrastructure is changing the Magnificent Seven38:36 Nvidia, Micron, semiconductors, and the risk of peak cycle earnings41:00 Why the biggest AI market stories could be scary for society43:37 AI disruption, labor markets, and the speed of technological change46:30 Measuring which jobs and companies are most exposed to AI automation49:00 Why AI cost structure may look more like Spotify than software51:13 The unresolved business model questions for LLMs and AI agents52:29 Why traditional value investing lost its edge56:03 Passive investing, book value, and the blame game in value investing58:13 Why rigid value investing is vulnerable to AI disruption01:00:58 How value investing can adapt to intangible assets and uncertainty01:02:21 Why any company can be a good investment at the right price01:04:57 Why investing mistakes and track records are harder to judge than they look

Edward Chancellor joins Kai Wu to discuss what financial history and capital cycle theory can teach investors about today’s AI boom. They explore why transformative technologies can still produce terrible investor returns, how overinvestment develops, where anti-bubbles may be forming, and what past episodes like the railway mania, the dot-com bubble, China’s investment boom and the post-2008 interest rate regime suggest about the risks and opportunities today.Guest links:Edward Chancellorhttps://www.edwardchancellor.com/Papers and articles discussed:Valuing AI: Extreme Bubble, New Golden Era, or Bothhttps://www.gmo.com/americas/research-library/valuing-ai-extreme-bubble-new-golden-era-or-both_viewpoints/Markets have poor scorecard for spotting AI losershttps://www.reuters.com/commentary/breakingviews/markets-have-poor-scorecard-spotting-ai-losers-2026-04-24/There’s no such thing as a good bubblehttps://www.reuters.com/commentary/breakingviews/theres-no-such-thing-good-bubble-2025-10-09/Big Booze can sweat off its multi-year hangoverhttps://www.reuters.com/commentary/breakingviews/big-booze-can-sweat-off-its-multi-year-hangover-2025-07-10/Topics covered:How capital cycle theory applies to the AI data center boomWhy railway mania, autos, aircraft and the dot-com bubble offer lessons for todayWhy markets often fund major technology transitions but fail to identify the winnersThe prisoner’s dilemma driving hyperscaler AI spendingWhether AI demand can justify the supply being builtHow GPU depreciation and AI capital spending may affect reported earningsWhy hallucinations and reliability may limit the total addressable market for large language modelsThe case for looking at AI anti-bubbles instead of shorting the bubble directlyWhy China shows that strong GDP growth does not guarantee strong shareholder returnsHow intangible capital, SaaS valuations and human capital fit into capital cycle analysisWhether bubbles can be good for society while still being bad for investorsWhy the long-term interest rate cycle may have changedThe role of gold in a world of expensive stocks, rising debt and vulnerable bondsTimestamps:00:00 Edward Chancellor on capital cycles, bubbles and AI04:42 Why the railway mania became a classic overinvestment cycle09:00 Why markets fund technology booms but often miss the winners13:19 The prisoner’s dilemma behind AI spending17:30 Will AI demand justify the supply being built20:00 How capital spending can inflate profits before the bust25:08 The AI Hindenburg moment and the limits of large language models30:55 Why AI hype may exceed the proven technology35:55 Why the anti-bubble may matter more than shorting AI40:00 The energy transition bubble and the opportunity in overlooked assets45:08 China’s lesson on GDP growth and shareholder returns49:27 Big Booze, GLP-1s and the Lindy effect54:23 Can intangible capital have its own capital cycle59:54 SaaS valuations and the index creation warning signal01:04:10 Why bubbles can help society but hurt investors01:09:09 Why long-term rates may be in a new multi-decade cycle01:14:07 Why Edward Chancellor still sees a role for gold