The Age of Abundant Intelligence and Scarce Bitcoin
AI erodes corporate moats; scarce assets like Bitcoin deserve portfolio exposure as a hedge.
View original →With over 30 years of Wall Street expertise in traditional finance and macroeconomics, I guide you through the evolving landscape of investments, from conventional assets to cryptocurrencies and AI-driven technologies.
Cite asCryptoQuant has tracked the directional calls on NVDA by @visserlabs on Substack since Sep 2024: not yet scored on NVDA (as of Nov 18, 2026).
2 NVDA calls on record, a thin history.
Relative view by default: fewer than 30 NVDA calls or days on record.
Sentiment timeline shows relative sentiment within this analyst's history. A perma-bull showing 8 bullish : 2 bearish in a bear market sets that as their baseline. If you notice any errors, claim @substack:visserlabs to submit corrections.
Hover the chart to read the thesis of that period · click a day to pin it
The Age of Abundant Intelligence and Scarce Bitcoin
AI erodes corporate moats; scarce assets like Bitcoin deserve portfolio exposure as a hedge.
View original →Build the Brain Before the Models Run Away
The most important AI story right now is speed. Every few months, the models get better. Then, suddenly, every few weeks, the models get better. Then one day you look around and realize that the thing many people dismissed as a chatbot has become a coding partner, a research assistant, a financial analyst, a tutor, a strategist, and a personal operating system. That is what the arrival of Fable 5 represents to me. The specific model matters less than the direction. The direction is obvious. The frontier is moving faster, the intelligence is becoming more usable, and the gap between the people using AI and the people ignoring AI is widening. There is a very important example hiding in plain sight. Since February 28, much of the market’s attention has been pulled toward the back-and-forth in the U.S.–Iran war, the risk around the Strait of Hormuz, and the possibility of another oil shock. That focus is understandable. But while everyone was watching geopolitics, the AI model layer kept accelerating. By my count, there have been roughly two dozen notable model releases, model-family launches, previews, or major capability rollouts since then. Anthropic alone shows the speed of the cycle: Claude Opus 4.7, then Opus 4.8, then Mythos Preview, and then Fable 5/Mythos 5. OpenAI released GPT-5.4 and GPT-5.5. DeepSeek released V4 Preview with a 1-million-token context window. Google, xAI, Mistral, Moonshot, MiniMax, Qwen, and NVIDIA all pushed forward as well. The point is simple: AI is not waiting for the world to calm down. That is the part I do not think most people understand yet. The risk is no longer that AI disappoints. The risk is that AI keeps improving while many people continue to treat it like a novelty. They try it once, ask it a question, get a generic answer, and decide they understand it. They use it like Google with a personality. Then they move on. Meanwhile, the models are moving from answering questions to doing work. That is the divide. Some people are still asking AI to write a paragraph. Others are asking AI to build tools. Some people are still debating whether AI is overhyped. Others are using it to build personal systems that make them smarter, faster, and more capable. That second group is beginning to separate. Last week, I saw something that made me more excited than almost anything I have seen in AI this year. I uploaded a project for subscribers (ai.22vresearch.com) showing how to build a Jensen Huang knowledge brain. The idea was simple: take transcripts, organize the source material, connect it to an AI workflow, use an API key, and build something that allows you to ask better questions of a focused body of knowledge. This was not meant to be a Silicon Valley engineering project. It was meant to be a bridge for people just like me with no prior coding experience. The goal was to take people who had mostly used AI as a chatbot and show them that they could build something. They could gather information. They could structure it. They could connect it. They could create a tool that reflected their own curiosity. The could feel empowered. The response I received in the days after was incredible. People who had never coded before were building. Adults who had been intimidated by AI were sending me messages saying they had done it. Some of them were proud in the way people are proud when they realize they can do something they thought belonged to another class of person. Then the next wave came, which was even better. Those same adults were showing it to their kids. That is when the whole thing clicked for me. This was not just a knowledge brain. This was agency. For the last year, AI has been discussed mostly through fear, market speculation, job loss, regulation, national security, and competition. All of those topics matter. I spend a lot of time on them. The macro implications are real. The investment implications are enormous. The national security implications are arriving faster than most people expected. Yet at the personal level, the most important question is much simpler. Do you feel more powerful because of AI, or do you feel more powerless? That is the entire game. The people who feel powerless will wait. They will watch the headlines. They will listen to the doomers. They will tell themselves they are too busy, too old, too nontechnical, too late, or too far removed from the center of the action. The people who feel powerful will build small things. That is how this starts. A knowledge brain does not need to begin with Jensen Huang. It can begin with anything you care about. Your favorite investor. Your favorite author. Your company’s internal documents. Your health research. Your industry. Your sales calls. Your podcast library. Your notes. Your family history. Your fantasy football preparation. That last one is where this gets fun, especially as we approach the beginning of training camps. As fantasy football season approaches, adults and kids are already starting to talk about building their own fantasy football brains. Imagine taking your favorite podcasts, transcripts, rankings, injury discussions, coaching comments, team previews, and draft strategy shows, then turning that material into your own AI-powered research assistant. Suddenly, AI is no longer abstract. It is no longer a scary headline about job replacement. It is no longer a debate about whether the models are conscious. It is no longer just a stock-market argument about capex, chips, power, and margins. It is no longer a way to secure a job. It is fun. It is a kid asking better draft questions. It is a parent and child building something together. It is a fantasy football brain that knows the voices you trust, the analysts you follow, the players you care about, and the strategies you want to test. That is the on-ramp. Most people do not become AI users because someone explains artificial general intelligence to them. They become AI users because they build one thing that matters to them. Once they build that one thing, their relationship with AI changes. The intimidation fades. The curiosity rises. The next project becomes easier. That is exactly why I built my AI subscriber paywall around three ideas: Signal, Alpha, and Agency. Signal means staying current. The AI world is moving too fast for people to follow casually. The headlines are noisy. The incentives are messy. The doomers are loud. The hype men are louder. One side wants you terrified. The other side wants you euphoric. Neither is useful by itself. Signal is about my YouTube channel filtering the noise and understanding the regime. What is actually changing? Which model release matters? Which infrastructure bottleneck matters? Which company comment matters? Which government action matters? Which part of the AI economy is accelerating, and which part is digesting the last wave of investment? If the models are speeding up, the first job is to stay oriented. Alpha means turning that signal into investment insight. AI is not just a technology story. It is a capital cycle. It is an infrastructure buildout. It is an energy story, a chip story, a data center story, a networking story, a software story, a security story, and eventually an application story. The winners will not be limited to the companies with the most famous chatbots. The buildout is much bigger than that. There will be beneficiaries across the stack, from power to chips to infrastructure to models to applications. The goal of the paywall is to take the signal and translate it into company-level opportunities, because investors need more than excitement. They need a map. Agency is the third piece, and right now it may be the most important. Agency means using AI yourself. It means prompts. Tools. Workflows. Experiments. Buildouts. Examples. Mistakes. Iteration. It means watching someone else do it and realizing you can do it too. The knowledge brain project was the clearest example yet. It gave people a path from passive user to active builder. It showed them that the API key, the transcripts, the code, and the AI assistant were pieces they could learn to assemble. Once they assembled them, they did not just have a tool. They had proof. Proof matters. Proof changes identity. A person who says “I do not know how to code” becomes a person who says “I built a knowledge brain.” A parent who worries their child is falling behind becomes a parent helping that child build a fantasy football research system. A subscriber who thought AI was something happening to the world begins to see AI as something they can use to shape their own world. That is the message I want people to hear. You do not need to become an AI researcher. You do not need to become a professional programmer. You do not need to understand every detail of model architecture, inference scaling, token economics, or frontier benchmarking. You need to start building. Start with something you love. Start with something familiar. Start with a small body of information and ask how AI can make it more useful. Build a brain around a person, a topic, a hobby, a market, a sport, a company, or a question you cannot stop thinking about. The first version can be messy. It should be messy. The goal is momentum. The models are going to keep improving. Fable 5 will lead to the next model, and the next model will lead to another step change after that. The people who wait for the perfect moment will discover that the perfect moment keeps moving away from them. The people who start now will compound. They will learn how to ask better questions. They will learn how to structure information. They will learn how to connect tools. They will learn how to separate good answers from lazy ones. They will learn how to turn AI from a chatbot into a collaborator. That is why I am so excited. The knowledge brain response showed me that the audience is ready. Adults are ready. Kids are ready. Investors are ready. Builders are ready. The curiosity is there. The only thing missing for many people is a path. That is what I want this platform to provide. Signal to know what matters. Alpha to understand who benefits. Agency to make sure you are not just watching the AI revolution from the sidelines. The world can always give you a reason to wait. The models will not. Build the brain before the gap gets wider.
View original →The Synthetic Reality Crisis
Why AI Is Forcing NFTs to Become Identity Infrastructure “But I think that may be easier to solve, strangely, than what happens to the human condition and humanity as a whole. And I think I’m also optimistic we’ll come up with new answers there. We do a lot of things today—from extreme sports to art—that aren’t necessarily directly to do with economic gain. So I think we will find meaning, and maybe there’ll be even more sophisticated versions of those activities.” — Demis Hassabis , Google DeepMind, World Economic Forum Annual Meeting 2026, Davos, Switzerland (January 20, 2026), on AI and human purpose Over the last two weeks, I’ve had multiple conversations centered on NFTs. As a macro investor, I’m always looking for how technology reshapes the world over time. While attending the recent Real Vision event in Miami, NFTs resurfaced for me in a meaningful way in what I think of as a moment of integration . Something that had been sitting quietly in the back of my mind suddenly snapped into focus, connecting multiple data points and bringing the mosaic together. I believe 2026 will mark the year of crypto utility . Stablecoins will continue to expand in real-world usage, driving network effects across the ecosystem. This acceleration will come from the growing interconnection between AI and payments, particularly as AI agents begin to transact on our behalf. Just as AI has turbocharged demand for NVIDIA, high-bandwidth memory, and even silver, it will do the same for stablecoin volumes and market capitalization. After two weeks of deep conversations with builders and investors in the NFT space, I now believe the timing is right there as well. I’ve mentioned tokenization often in recent months but have largely avoided NFTs, until now. That changed over the last two weeks. This paper explores AI’s disruptive impact on trust, the growing drift in human identity, and why NFTs emerge as a macro theme in that environment helping address one of the central questions of the AI disruption: what role do humans play in a world where traditional jobs may no longer define purpose? One final thought. I had a long conversation with Vi Powils and Sweetpea from World of Women during this trip. Their commitment to the space and enthusiasm, along with conversations with longtime friends and new ones at the conference, helped shape this moment for me. The entrepreneurs in NFTs who survived the bubble of the last few years remind me of the builders who emerged after the dot-com collapse. I’m grateful for those still building through the noise. NFTs will have a meaningful impact on society in the years ahead due to AI. The foundation is being laid now. The Dual Crisis: Truth and Purpose You’ve felt it. The scroll that used to be mindless now carries a low-grade dread. A video that feels satisfying, was it real? A news clip that seems damning, is it synthetic? A voice note from someone you love, do you need to double-check it’s actually them? And beneath that, a different unease. You read that lawyers are being replaced. Then radiologists. Then designers, analysts, writers and soon all knowledge workers. Each headline lands in the same place: the thing you spent years becoming might not belong to you much longer. The younger you are, the harder it is to ignore. These anxieties are usually discussed separately. They shouldn’t be. They are the same crisis, arriving at the same time, moving faster than we can metabolize. We are entering a period of human history defined not by a single disruption, but by the convergence of two. The first is a crisis of truth. Generative AI has severed the relationship between digital representation and physical reality. Images no longer guarantee light was captured. Voices no longer guarantee a human spoke. Video no longer guarantees presence. Deepfakes, synthetic media, and infinite content generation have collapsed the default trust assumptions that underpinned the internet for decades. The second is a crisis of purpose. AI is not merely automating repetitive labor; it is outperforming humans at cognitive labor itself. Professions that once anchored identity, law, medicine, finance, journalism, design are being systematically devalued. The question “What do you do?” is losing its power to answer the deeper question: “Who are you?” Together, these forces create something unprecedented: an epistemic and identity vacuum. A world in which neither our experiences nor our contributions can be reliably verified, and in which the work we once used to define ourselves no longer belongs uniquely to us. This is not a technological inconvenience. It is an existential challenge. From Occupational Identity to the Identity Vacuum For most of human history, identity and labor were inseparable. Surnames encoded vocation directly into selfhood, Miller, Smith, Baker, Cooper. Visser means fisherman in Dutch. To know who someone was, you only needed to know what they did. This fusion persisted into the modern era, abstracted but intact. “I’m a doctor.” “I’m a lawyer.” “I’m an engineer.” Résumés, credentials, and LinkedIn profiles became the architecture of the professional self, titles stacked like achievements, each reinforcing identity. AI threatens to demolish this structure entirely. When an algorithm diagnoses disease more accurately than a physician, or writes legal briefs faster than an associate, or recognizes market patterns better than a seasoned analyst, the damage is not only economic. It is psychological. The identity constructed around expertise collapses. Humans have never navigated a transition where cognitive labor itself becomes non-sovereign. When vocational identity dissolves, identity does not disappear. It migrates. Identity Beyond Production: The Experiential Self Strip away job titles and something fundamental remains: experience. “I ran the Boston Marathon.” “I was at Woodstock.” “I waited 108 years for the Cubs to win.” “I was there when history happened.” These are not claims of productivity. They are claims of presence. They derive meaning from having been somewhere, witnessed something, endured uncertainty, and participated in moments that unfolded over time, often in community, often through struggle. This experiential identity has always existed alongside professional identity. But as AI devalues what we produce, it elevates what we experience and witness . The problem is that AI undermines the ability to verify those experiences at the exact moment they become most important. The Epistemic Void: When Experience Becomes Unprovable Consider a grandmother receiving a phone call from her “grandson” asking for emergency bail money. The voice is perfect. The urgency is real. The call is fake. Or a journalist publishing footage of a war crime, only to have it dismissed as AI-generated propaganda. Or a lifelong fan whose photos from a historic game are waved away as synthetic recreations. These are not edge cases. They represent the collapse of the verification layer of society. We now live in what can be described as an Epistemic Void: a world where the provenance of any digital artifact is unknowable without cryptographic proof. This void affects truth and identity simultaneously. The grandmother and the displaced professional face the same problem: the systems that verified who we are and what we experienced no longer work. Photos can be fabricated. Credentials can be simulated. Community participation can be spoofed by bots. When identity cannot be verified, anxiety follows. Humans require ontological security, the sense that reality is stable and that our experiences are real, recognized, and acknowledged by others. Without this, social trust frays. From Wealth Classes to Communities of Meaning As vocational identity thinned in the pre-AI era, society attempted a substitute. Community fractured into classes of wealth rather than communities of meaning. Status became economic. Belonging was purchased rather than earned. Zip codes, access, and consumption replaced shared experience. These structures were fragile. They are easily mimicked by bots, easily arbitraged by capital, and easily hollowed out by abundance. AI accelerates their collapse. When status goods can be generated, copied, or simulated infinitely, their signaling power evaporates. In response, a new scarcity emerges: verified authentic experience. Identity migrates away from what we own and toward what we have witnessed together . The Scarcity of Authentic Experience This inversion is central. In the analog world, experience was common and records were scarce. In the early digital world, records became abundant but still traceable to reality. In the AI world, synthetic records become infinite, while verified authentic records become scarce. Anyone can claim anything. Anyone can generate evidence. The only thing that matters is what can be cryptographically anchored to reality. This is where belief, time, and identity converge. The Damien Hirst Experiment: Belief Before Proof In 2021, Damien Hirst conducted one of the most revealing experiments of the digital age, though it was widely misunderstood as an art-world stunt. The Currency forced collectors to choose between a physical artwork and a corresponding digital token. Whichever they selected, the other would be destroyed. There was no right answer, no historical precedent, only belief and a deadline. The experiment was not about art markets. It was about trust under uncertainty. Collectors were forced to decide what they believed would matter more in the future: physical presence or digital permanence. Time enforced conviction. Belief preceded validation. In a world where AI makes representation cheap and infinitely reproducible, Hirst’s experiment previewed the coming identity dilemma. When truth cannot be inferred visually, value migrates toward what can be proven and what is believed early enough to endure uncertainty. NFTs as Identity Infrastructure This reframing clarifies the role of NFTs. Stripped of speculation, an NFT is not a collectible. It is an immutable witness statement. When properly architected, NFTs provide cryptographic answers to identity questions that AI cannot counterfeit: Proof of Presence — verifiable evidence that a human was physically present at a specific place and time Proof of Participation — confirmation that an experience was completed, not simulated Proof of Belonging — evidence of sustained community membership earned over time Proof of Witness — an immutable anchor tying memory to a real historical moment NFTs do not create experience. They verify that it happened. In a synthetic reality, verification becomes identity. Utility as the Transmission Mechanism If identity is the why behind NFTs, utility is the how . Infrastructure does not spread through belief alone; it spreads by embedding itself into everyday function. This was the lesson of the post-dotcom era. The internet did not win because people philosophically agreed with it. It won because email worked, commerce worked, and social connection compounded. NFTs are now undergoing a similar transition. What began as speculative representation is evolving into functional identity primitives: access credentials, participation proofs, community passports, and continuity layers between physical and digital life. AI is the forcing function. As synthetic media erodes trust, systems that can verify presence, participation, and belonging acquire immediate utility, not as investments, but as tools. The growth of NFTs will not be driven by hype cycles, but by repeated use in moments that matter: entry, access, memory, and community. Utility is not the thesis. It is the delivery system that makes the thesis unavoidable. The Alchemy of Shared Suffering No domain illustrates this more clearly than sports. A fan’s identity is not forged at the trophy ceremony. It is forged in years, often decades in my case, of shared suffering. Losses. Near-misses. Enduring hope without certainty. The joy of the win exists only because of the history of loss. In a world where success can be algorithmically generated, the Human Premium lies in uncompensated commitment to a journey whose outcome was unknown. Being there matters. AI cannot endure. It cannot belong. It cannot suffer through time. The Human Premium As AI capabilities expand, human experiences do not become less valuable. They become differently valuable. AI can generate output. AI can simulate knowledge. AI can fabricate evidence. But it cannot be present. It cannot witness time as a human. It cannot form authentic communal memory. What remains valuable is not what we produce, but where we were, what we witnessed, and who we became through presence and participation. NFTs, properly understood, are the verification layer of that humanity. They are not surveillance. They are sovereignty, the ability to prove, without permission: I was there. I endured. I belong. The Epistemic Void is real. The identity crisis is accelerating. In a world where machines can fake anything and automate everything, the only thing that cannot be faked is being human.
View original →Forget the PowerPoints Decks: Podcasts Are the New Lab for AI Thinkers
School was torture for me. Not because I couldn’t learn. My grades were good enough to keep my parents from being on top of me. It probably helped that my entrepreneurial father had dropped out of high school. I honestly just couldn’t force myself to sit through content that bored me senseless. While teachers droned on about topics that felt dead on arrival, my mind was elsewhere, restless, searching for something that would actually capture my attention. Before teachers get mad at me, your job was not to entertain me or try to keep my attention so I get it, it’s me, not you. It wasn’t until years later that I figured out the pattern: I’m an insatiable learner when it comes to things that interest me, and I can only handle content that doesn’t bore me. School was the worst possible combination of both with forced consumption of material that failed to earn my attention and teachers who were not selected on their TED Talk abilities. What I’ve realized since then is that my experience wasn’t unique, it was diagnostic. The problem wasn’t me or school exactly; it was the fundamental mismatch between how insight actually happens and how institutions try to package it. Nowhere is this more obvious today in this fast-moving AI world than in financial markets. PowerPoint presentations are a microcosm of this problem and basically like going back to school for me. I stopped attending meetings many years ago because of the 100-page decks people handed out to show me their views. It is when they would say, “go to page 85” and they knew what was on the page that I said, enough is enough. This is the institutional equivalent of that boring classroom lecture: polished, structured, static, and often dead by the time it reaches you. I just want to talk to someone whose view evolves every day. Not change, evolve! Now fast forward to today. AI is the greatest class I have ever attended. Each day things have changed so much, you can’t keep up. This is why I feel bad for sell-side research. Having worked at Morgan Stanley more than 20 years ago, I still remember how long it took to go through the bureaucracy of getting anything out. I am sure it has improved (maybe?) but there is no way AI waits around for legal and compliance to say, “all good.” And for anyone reading books on AI, oh my god, you might as well go get your money back now. Now you know why in my weekly video of the week’s events for macro, AI and crypto, it is filled with news from AI and what really consumes most of my research time, podcasts. The rise of podcasts was a gift for me and my limited attention span. Podcasts became my real education and I get to choose the teachers whom I could name for you one by one. I judge my podcasts based on what I called from my prior podcast as “Green Marbles.” This was something I would describe to people as data points (green marbles) that stood out to me from the other marbles (all the other colors) that connected prior green marbles to one another. It is a filtering process of puzzle pieces in the bigger picture. As an example, every time Marc Andreessen speaks, the green marble connections ring in my head like a pinball machine. I have learned, for what it is worth, that if it is difficult to listen to someone at 1.75x speed, they usually can do that to me. Marc is one of them and so is Dylan Patel. Basically, again, it is the opposite of a PowerPoint presentation where they know their topic so well that they are able to speak about it without thinking. The Deck: Yesterday’s News in Tomorrow’s Market The problem for me using AI sell-side research as an example is it is built for a world that no longer exists. Analysts spend weeks building models, crafting narratives, and getting compliance approval only to deliver conclusions that were already outdated the moment research began. The format rewards certainty over curiosity, polish over process. By the time a deck hits your inbox, the thinking has stopped. What you’re reading is a performance, not an exploration. This was always a limitation, but it’s become fatal in the age of AI. The exponential pace of change has completely broken the traditional research cycle. NVIDIA’s data center revenue grew tenfold in less than two years. Hyperscalers were expected to have capex numbers around 300 billion for this year and numbers will end much higher than that. New model architectures emerge monthly, reshaping the economics of compute. What was true Monday is obsolete by Friday. For analysts to stay relevant, they’d need to revise their decks weekly, sometimes daily. But compliance reviews don’t run on that schedule. Legal exposure creates conservatism. The institutional apparatus that gives research its authority also makes it impossibly slow. I have noticed this more and more in my conversations about AI demand where funds will tell me “experts” see an oversupply coming soon. I always say the same thing, it isn’t about supply, it is about demand and demand changes by the podcast. The Regulatory Straitjacket Here’s the thing nobody talks about: sell-side research literally cannot say the most interesting things. Reg FD and securities laws create a chilling effect on forward-looking speculation. Analysts can’t share hunches, patterns they’re sensing, or intuitions not yet backed by hard data. They can’t say “I don’t have proof, but I’m seeing momentum build in this area.” Everything must be fact-based, defensible, conservative. Multiple reviewers strip out edge, nuance, and speculation precisely the elements that generate actual insight. By the time research clears compliance, it’s been sanitized into institutional speech that avoids saying anything risky or interesting. Combine all of that with exponential demand and you have a mismatch. Podcasts operate in a completely different universe. They’re not investment advice or formal research, they’re conversation. That distinction creates permission for genuine exploration. An expert can say “this is speculation, but here’s my intuition...” or “the data doesn’t exist yet, but here’s how I’m thinking about it.” These statements would never survive compliance review, yet they often contain the seeds of genuine insight. When Sam Altman recently said “capacity overhang” to describe the models OpenAI had but could not release I remember pausing the podcast on my walk in Williamsburg and quickly typing into ChatGPT to start a dialogue looking for other examples of this to confirm. 30 minutes later, I had multiple data points to use in my weekly video. Why Conversation Works: The Neuroscience of Real-Time Thinking When someone speaks in an interview which is fast-moving in long form, they’re not reciting conclusions, they’re generating ideas in real time. The brain activates different networks during improvisational speech: broader semantic memory, associative connections across domains, spontaneous pattern-matching. Speaking forces you to retrieve and connect knowledge on demand, often in combinations you’ve never attempted before. As someone who does a one-take weekly video for 45 minutes with no script, I can tell you from experience, I don’t know what is going to come out until it does. This is why experts often “hear themselves think” during conversations. The act of verbalization becomes a feedback loop. You start answering a question with one framework in mind, but as you speak, you notice gaps or contradictions, forcing real-time adjustment. You articulate something you thought you understood, only to realize through explanation that your understanding was incomplete. A skilled interviewer amplifies this effect. They push into unexplored territory, ask questions that don’t have rehearsed answers, surface contradictions, and force synthesis. The best podcast moments happen when an expert is genuinely surprised by their own answer, when the conversation pulls together a pattern they’d never consciously articulated. This is cognition in action. Not the final product of thinking, but the actual process of discovery. For listeners trying to navigate exponential change, watching that process unfold is far more valuable than receiving polished conclusions. Authenticity vs. Performance Sell-side decks are performances of certainty. Analysts must project confidence even when the future is genuinely unclear. The format doesn’t permit “I don’t know” or “I’m working through this” or “this could go either way.” Every ambiguity gets resolved into a definitive view because institutional credibility requires it. Podcasts reward authenticity. The best moments are when someone admits uncertainty, acknowledges the limits of their knowledge, or revises their thinking mid-conversation. This isn’t weakness, it’s signal. In markets where overconfidence is endemic and humility is rare, an expert who says “I don’t know yet, but here’s how I’m approaching the question” stands out. You learn not just what they think, but how they think, which is the only thing that stays relevant when conditions change faster than analysis can keep pace. Speed, Iteration, and the Living Record Podcasts solve the velocity problem not through speed alone, but through mode. When a major AI announcement drops, an expert can appear on a podcast within days, processing implications in real time, raising questions that don’t yet have answers. The conversation becomes a living record of thought adapting to new information. I always search by news if ithere is something specific I want to learn about. More importantly, experts can iterate across multiple conversations. Each podcast refines the thesis, incorporating feedback and new data. This happens at conversational speed, days, not months. No compliance reviews, no institutional approvals. Just intelligence engaging with complexity in real time. The format also creates accountability of a different kind. In sell-side research, analysts are judged on whether their price targets hit. This encourages either conservatism (stick to consensus) or sensationalism (make bold calls that get noticed). In podcasts, credibility depends on reasoning quality. An expert who thinks clearly, acknowledges counterarguments, and demonstrates humility maintains credibility even when specific predictions miss. And again, in a world of exponential change, price targets are becoming more random and another forced event from the industrial age. Dialogue Is the New Alpha As AI automates quantitative analysis and pattern recognition, human advantage concentrates in domains where machines struggle: contextual understanding, analogical reasoning, real-time sense-making of ambiguous situations. These are fundamentally conversational capacities. They emerge through dialogue, not datasets. In markets where information is commoditized, the scarce resource is wisdom, the ability to distinguish signal from noise, to reason under uncertainty, to see patterns emerging from chaos. Wisdom cannot be programmed into a deck or automated by an algorithm. It’s situational, contextual, and emerges through articulation. Podcasts are intelligence systems for an age of acceleration. They’re messy, unpolished, often meandering. But in that messiness lies their power: the willingness to discover rather than recite, to stay alive to possibility, to think out loud. They capture thought in its natural habitat, alive, social, perpetually unfinished. Finding My Classroom I spent years thinking I was a bad student. Turns out I was just in the wrong classroom. Your child may be the same, especially in this fast-moving world of social media and AI. The traditional model, static content, predetermined conclusions, one-way transmission was designed for an industrial age that moved slowly and valued standardization over curiosity. Podcasts represent something fundamentally different: learning optimized for discovery rather than distribution. When I’m listening to experts think through AI infrastructure bottlenecks or crypto protocol design, I’m not receiving conclusions. I’m watching someone navigate uncertainty in real time, which teaches me how to do the same. If I want to expand on something I heard, I switch to ChatGPT and have a conversation where the context window is saved in case I want to go back to it later. Think of it as asking the smartest person you know about the topic, so you learn it in more depth. That restless energy I had in school? It was never about inability to focus. It was about refusing to waste attention on dead content. Podcasts earned my attention the way school never could, by staying alive, by revealing thinking rather than reciting conclusions, by treating listeners as intelligent participants rather than passive recipients. In a world of exponential change, the last competitive advantage is the willingness to keep thinking, keep questioning, keep adapting. As I share podcasts with people, I regularly hear, I don’t have the time. If that person is you, ask yourself this question- “Did you sit in a meeting today that was a waste of time?” If you say no, you are lying unless you had no meetings that day. Podcasts are simply the medium that best captures and transmits that adaptive intelligence. The future belongs to dialogue. The classroom I was looking for all along was just waiting for the technology to catch up.
View original →Edison’s Lesson for the AI Age: Why the Capex Boom Isn’t a Bubble
AI capex is not a bubble; hyperscalers like Amazon will capture application value.
View original →