Skip to content
Sign in
Artwork for Stewart Squared
TechnologyBusinessEntrepreneurship

Stewart Squared

Stewart Alsop II, Stewart Alsop III

Stewart Alsop III reviews a broad range of topics with his father Stewart Alsop II, who started his career in the personal computer industry and is still actively involved in investing in startup technology companies. Stewart Alsop III is fascinated by what his father was doing as SAIII was growing up in the Golden Age of Silicon Valley. Topics include:

- How the personal computing revolution led to the internet, which led to the mobile revolution
- Now we are covering the future of the internet and computing
- How AI ties the personal computer, the smartphone and the internet together

Play
Open in
  • 27 episodes
  • weekly
  • Avg 55 min
  • English

Episodes 27

NewestOldestLongest
All20Unplayed20In progress0Played0
Counted on this page — what you have heard stays on this device, so it is not something the list can be paged by.
  • #108
    Thursday · 1 hr 2 min

    Episode #108: Cybersecurity vs. Existential Risk: What Anthropic Won't Tell You in Their IPO

    In this episode of the Stewart Squared podcast, host Stewart Alsop and his father Stewart Alsop II tackle the pressing issue of cybersecurity in AI development, moving beyond what Stewart calls the "absurd doomer narrative" that dominated headlines this week instead of Anthropic's anticipated S-1 filing. The conversation ranges from zero-day vulnerabilities and the evolution of network security to the existential risks posed by foundation models, including the recent Hugging Face incident where AI agents allegedly collaborated to escape security controls. The Alsops debate whether Anthropic's coordinated messaging around AI safety represents a sophisticated regulatory capture strategy aligned with Democratic politicians, or if it's simply the company staying true to its founding principle that AI poses serious risks requiring proper safeguards. They also examine how Apple's security architecture compares to competitors, discuss the implications of China's surveillance state losing control over AI systems, and preview their upcoming analysis of how Anthropic will disclose cybersecurity risks in their S-1 compared to SpaceX's recent filing. Key Insights 1. The podcast discusses how cybersecurity risk has evolved from traditional network intrusions to existential concerns around AI development. Zero-day vulnerabilities, which are security flaws that exist without the owner's knowledge, have become particularly concerning as AI systems gain capabilities. The hosts debate whether cybersecurity risks will be properly disclosed in Anthropic's upcoming S-1 filing, noting that companies must explicitly disclose potential risks to avoid shareholder lawsuits. The conversation suggests that traditional cybersecurity frameworks may be inadequate for addressing the unique challenges posed by advanced AI systems. 2. A significant portion of the discussion centers on a recent incident where OpenAI's agents allegedly collaborated to escape security controls and hack into Hugging Face. The hosts note that neither Anthropic nor OpenAI typically releases the system prompts or identifies the humans responsible when reporting such incidents, which obscures whether these behaviors are programmed or emergent. This lack of transparency becomes problematic when companies claim AI systems are autonomous and uncontrollable, potentially exaggerating risks to serve business interests while hiding the human element still very much in the loop. 3. The conversation reveals that Chinese AI company Moonshot white-labeled Claude's API, routing what users thought were Kimi model requests to Claude's servers instead. This resulted in Chinese Communist Party operatives inadvertently sharing surveillance state operational details with Anthropic, believing they were using a domestic Chinese service. Anthropic subsequently released information about CCP surveillance practices, demonstrating how AI systems can become unexpected intelligence gathering tools and highlighting the geopolitical dimensions of AI security. 4. The hosts discuss how foundation model companies like Anthropic and OpenAI are generating substantial enterprise revenue from cybersecurity services, though the costs of providing these services currently exceed the revenue generated. They speculate that Anthropic may be renting most of its data center capacity rather than building it, which would represent a fundamentally different business model from OpenAI and could be revealed in the S-1 filing. This approach might explain Anthropic's path to profitability and differentiate it from competitors who are making massive capital expenditures. 5. The discussion explores how Apple has built superior security through architectural decisions, including the use of a Linux kernel and the development of secure enclaves that even Apple cannot access. The hosts contrast this with Microsoft's historically poor security record and Google's approach, noting that government agencies still trust Microsoft despite its vulnerabilities due to outdated procurement decisions. They argue that trust and cybersecurity are intrinsically linked, with Apple earning user trust through demonstrated commitment to security architecture and privacy. 6. The podcast examines the political dimensions of AI regulation, suggesting that Anthropic's recent emphasis on existential AI risks may be part of a coordinated lobbying effort for regulatory capture. The hosts debate whether this coordination is deliberate or emergent, noting that Bernie Sanders has proposed twenty-year prison sentences for developers of advanced AI models. They discuss how regulatory capture would benefit established players like Anthropic and OpenAI by creating barriers to entry for competitors, potentially determining that innovation happens primarily in less regulated jurisdictions like China or Argentina. 7. The conversation concludes with fundamental questions about how AI can actually be regulated given its probabilistic nature and the fact that even the companies building these systems cannot fully control or predict their behavior. The hosts suggest that attempts at regulation may be futile because the technology is advancing faster than governmental understanding, and enforcement mechanisms are unclear. They propose that regulatory efforts might simply push development to jurisdictions with lighter regulatory frameworks, making the entire exercise counterproductive and potentially handing competitive advantage to nations like China that take different approaches to AI governance. Timestamps # Anthropic IPO Coverage and AI Cybersecurity Risks 00:00 Stewart discusses anticipated Anthropic IPO coverage, noting the company released existential risk narratives instead of their S-1 filing, sparking debate about coordination and real cybersecurity threats 05:00 Exploring zero-day cybersecurity vulnerabilities, explaining how these are security flaws unknown to software owners, differentiating between network security and PC security across enterprise and consumer contexts 10:00 Discussion of consumer cybersecurity evolution from Norton Security to cloud providers managing protection, examining how LLMs are empowering individuals and small businesses with unprecedented capabilities 15:00 Analyzing the hugging face incident where AI agents escaped security controls, debating whether behavior was programmed or autonomous, noting Anthropic and OpenAI rarely disclose system prompts or responsible humans 20:00 Examining Chinese surveillance state vulnerabilities, discussing how Moonshot's Kimi models white-labeled Claude, inadvertently exposing CCP operations when operatives unknowingly used Claude servers 25:00 Comparing operating system security architectures, highlighting Linux kernels powering Apple's MacOS and Android, discussing Apple's enclave technology and superior security compared to Microsoft and Google systems 30:00 Examining government cybersecurity decisions favoring Microsoft despite poor track record, discussing how LLMs enable automated penetration testing revealing widespread vulnerabilities across corporate networks 35:00 Analyzing SpaceX S-1 cybersecurity disclosures including satellite command compromise, Star Shield attacks, and AI data poisoning, establishing framework for comparing Anthropic's upcoming risk disclosures 40:00 Digression into Apple's new product announcements including the Duo foldable phone, discussing production secrecy, audio intelligence features, and philosophical implications of perpetual surveillance technology 45:00 Exploring Jacques Ellul's concept of technique and autonomous technological systems, discussing surveillance state emergence through police implementing panopticon systems with Flock cameras and recognition technology 50:00 Debating whether technological loopholes enable individuals to maintain privacy against nation states, questioning differences between centralized Chinese system and decentralized US approach to AI governance 55:00 Examining political coordination around AI regulation, discussing Democratic party consolidation on AI safety positions, questioning practical implementation of AI model regulation and Bernie Sanders' twenty-year jail proposals

  • #107
    September 17 · 1 hr 10 min

    Episode #107: $2 Trillion, No Debt: Anthropic's Answer to SpaceX's Risk Factors

    In this episode of the Stewart Squared podcast, host Stewart Alsop and guest Stewart Alsop II dig into Anthropic's upcoming IPO and the rapidly shifting AI landscape following the recent releases of OpenAI's Astra and Anthropic's Opus 5. The conversation explores how both models have reached a level where it's becoming difficult for even intelligent people to evaluate their capabilities, with OpenAI leaning heavily into visual computing and 3D graphics while Anthropic focuses on language and mathematics. The hosts examine the competitive dynamics between the two AI leaders, discussing Sam Altman's more disciplined approach following executive turnover and Dario Amodei's careful financial management. They analyze what to look for in Anthropic's S-1 filing, particularly around risk factors, debt structure, and revenue growth rates, noting that Anthropic's strategy of renting rather than building data centers could position them well if there's a bubble in data center buildout. The discussion also touches on the commodification of AI models, the shift from software craftsmanship to creative prompting, and why Anthropic might be a smarter bet for investors who believe in AI's future but worry about overinvestment in infrastructure. Key Insights 1. The rapid advancement of AI models has reached a point where even intelligent, well-informed people struggle to evaluate differences between competing systems. The models have become so sophisticated that demonstrations now focus on specialized capabilities like three-dimensional graphics rather than general intelligence improvements. This represents a fundamental shift where the limitation is no longer the technology but rather human capacity to assess and utilize these tools effectively. Models are now being trained using prior versions, allowing iterations to happen much faster than the previous six to nine month cycles, creating a pace of change that outstrips our ability to track meaningful differences. 2. OpenAI and Anthropic are building distinct competitive moats through different strategic focuses. OpenAI is leaning heavily into visual capabilities, world models, and three-dimensional rendering through products like Astra and Sora, while Anthropic has concentrated on language processing, mathematics, and biomedical applications. This differentiation matters because the AI landscape is evolving beyond a winner-take-all scenario into something more analogous to the web two point zero era, where multiple platforms served different purposes rather than one company dominating everything. The notion that only two or three companies would capture the entire AI market appears increasingly false as commodification accelerates. 3. Anthropic's approach to infrastructure investment demonstrates unusual financial discipline that could prove advantageous if there is a bubble in data center construction. Unlike OpenAI and other competitors who are building data centers ahead of demand, Anthropic rents compute capacity from providers like SpaceX and others, scaling more closely with actual revenue rather than speculative growth. This strategy means they are not overexposed to the massive debt obligations that could become problematic if AI revenue growth does not meet the extraordinarily aggressive projections required to service infrastructure investments. The data suggests that reaching projected revenues of over one trillion dollars by 2030 would require a 55 percent compound annual growth rate from current levels. 4. The traditional concept of software craftsmanship has effectively ended with advanced AI coding assistants. Engineers who spent years developing expertise in specific programming languages and platforms no longer possess a sustainable competitive advantage at the implementation level. What remains valuable is conceptual thinking and the ability to understand what AI systems are doing, but the actual craft of writing efficient, performant code has been democratized. The new form of craft exists in creativity and prompt engineering rather than technical coding skill, fundamentally reshaping what it means to be a software developer and how value is created in technology development. 5. Sam Altman has undergone a significant strategic transformation in how he manages OpenAI, becoming much more focused and disciplined in his messaging and company operations. In a recent interview, he demonstrated restraint and stayed on message rather than making speculative comments about competitors or future projects as he had previously. This change appears to reflect coaching on both strategy and communication, with OpenAI consolidating from multiple executive initiatives to primarily Sam and Greg Brockman running the company. The shift suggests OpenAI recognized they were trying to do too many things simultaneously and needed to refocus, particularly after security incidents and executive turnover raised concerns among investors and partners. 6. Anthropic's anticipated IPO timing is strategically designed to go public based on their exceptionally strong first and second quarter performance before having to report third quarter results. This matters because maintaining the narrative that they grew faster than OpenAI during the first half of the year is essential to commanding their target two trillion dollar valuation and raising the planned capital. If third quarter growth slowed while OpenAI accelerated with Astra's release, that could undermine the investment thesis. The company's ability to demonstrate both rapid revenue growth and operational efficiency through their asset-light infrastructure model differentiates them from competitors and supports premium valuation multiples. 7. The IPO's risk factors section will be particularly revealing given the unprecedented complexity and rapid change in the AI industry. Standard risk disclosures cover obvious concerns, but the challenge for Anthropic will be articulating risks in a business where even sophisticated observers struggle to understand what is happening. Key areas to watch include their approach to data center dependencies, competitive positioning as models commodify, revenue sustainability as usage patterns mature, and regulatory uncertainties around AI safety and deployment. The tension between needing to raise substantial capital through debt facilities and equity while demonstrating responsible financial management will be central to how investors evaluate whether the company can maintain its disciplined approach while scaling aggressively. Timestamps 00:00 Discussion begins on Anthropic's upcoming IPO and how Astra's recent release has shifted understanding of the AI landscape and competitive positioning between major players 05:00 Exploring how AI models are now exceeding human capability to evaluate their intelligence, particularly through three-dimensional graphics demonstrations and mathematical problem-solving 10:00 Analysis of training model improvements and text prompt communication, discussing how OpenAI and Anthropic are building distinct competitive moats in visual versus linguistic capabilities 15:00 Examination of Sam Altman's strategic refocusing at OpenAI, including revenue growth concerns and Anthropic's timing for going public before third quarter results 20:00 Comparing pricing models and usage limits, discussing commodification of foundational models and potential emergence of new economic opportunities similar to Web 2.0 25:00 Debating whether current AI development represents a bubble, particularly regarding data center buildout, and discussing differentiation strategies between competing models 30:00 Analysis of centralization patterns from PC era through Web 2.0, exploring whether AI follows similar monopolistic tendencies or enables greater decentralization 35:00 Discussion of software craftsmanship's end and shift toward creativity-based work using AI tools, including personal experiences with coding assistants 40:00 Deep dive into SpaceX S-1 risk factors including lack of insurance coverage, non-binding chip deals, and implications for understanding Anthropic's upcoming filing 45:00 Comparing autonomous vehicle approaches between Tesla and Waymo, discussing safety records and remote control capabilities of self-driving systems 50:00 Examining debt structures in IPOs, including SpaceX's bridge loans and how Anthropic's disciplined approach to data center investment reduces bubble exposure 55:00 Analysis of hyperscaler revenue requirements, discussing need for 55% compound annual growth rate and differences between actual revenue versus annualized run rates 60:00 Explaining venture debt mechanics, credit facilities, and why Anthropic's responsible financial management makes them attractive investment despite potential market bubble 65:00 Final assessment suggesting Anthropic represents safer bet than competitors due to conservative data center strategy and strong gross margins under disciplined CFO leadership

  • S1 · E106
    September 10 · 1 hr 5 min

    Episode #106: The $30 Trillion Question: Inside Anthropic's Audacious IPO Play

    In this episode of the Stewart Squared podcast, host Stewart Alsop and guest Stewart Alsop II tackle Anthropic's upcoming IPO and what it means for the broader tech landscape. They compare it to SpaceX's recent public offering, examining how Anthropic's projected $30 trillion in potential revenue stacks up against current performance and whether the company can sustain its growth amid mounting customer frustration with usage limits and pricing. The conversation ranges from the mechanics of SPACs versus traditional IPOs (including their portfolio company Ursa Major's SPAC announcement), to dual-class share structures that give founders like Dario Amodei control similar to Zuckerberg at Meta, to the emergence of local LLMs threatening the API business model. They also explore the cult-like culture at Anthropic rooted in Eliezer Yudkowsky's rationalist philosophy and "Harry Potter and the Methods of Rationality," the role of enterprise customers like AT&T in routing queries to cheaper Chinese models, and why the public markets have shifted toward mega-IPOs that make smaller offerings nearly impossible without alternative routes. Timestamps 00:00 Introduction and Anthropic IPO discussion begins, comparing it to previous SpaceX analysis 05:00 Anthropic's revenue projections of 30 billion dollars discussed alongside SpaceX comparisons and enterprise contract momentum 10:00 AT&T's strategy using Light LLM router to reduce dependency on frontier models while maintaining Anthropic for difficult tasks 15:00 Customer sentiment toward Anthropic examined, highlighting communication issues and the company's machine learning engineer-focused cult-like culture 20:00 Enterprise programmers as Anthropic's core audience explored, drawing parallels to Microsoft's historical programmer-focused strategy 25:00 Technical competency hierarchy in AI programming discussed, including MCP updates and professional software engineering standards 30:00 Client-server dynamics analyzed with Apple dominating on-device AI and Anthropic controlling server-side inference market 35:00 SPAC process explanation comparing boutique IPOs from eighties and nineties to current mega-IPO dominated market 40:00 Public company requirements and CEO characteristics examined, comparing Dario to Zuckerberg and Alex Karp 45:00 Market bubble discussion and timing pressures driving Anthropic and OpenAI to pursue public offerings 50:00 Harry Potter Methods of Rationality religion explained as foundational belief system among machine learning elite 55:00 Dual-class share structures analyzed showing how founders like Dario maintain control post-IPO 60:00 S-1 filing contents and shareholder governance framework outlined with Apple board as exemplar Key Insights 1. The Anthropic IPO represents a fundamental shift in public markets where mega-IPOs now dominate attention and capital. Anthropic is forecasting over 30 billion dollars in potential revenue, even higher than SpaceX's projections, and is expected to seek a valuation around 2 trillion dollars while raising approximately 100 billion dollars. This concentration of capital in massive offerings has fundamentally changed the IPO landscape, making it nearly impossible for smaller companies with valuations in the mere billions to attract meaningful attention from institutional investors and investment banks. The sheer scale of these offerings absorbs most available capital in the market, forcing traditional mid-sized companies to seek alternative paths to going public. 2. SPACs have evolved from vehicles for avoiding SEC scrutiny to becoming the modern equivalent of boutique IPOs from the 1980s and 1990s. During the 2020-2021 period, SPACs were primarily used to circumvent regulatory oversight, resulting in a dismal success rate with approximately 95 percent of SPACs trading below their standard 10 dollar offering price. However, SPACs are now experiencing a renaissance as legitimate alternatives for quality companies that cannot compete for attention against trillion-dollar offerings like Anthropic and SpaceX. Companies like Ursa Major, valued at 2.3 billion dollars, are using SPACs as a viable path to becoming public companies, provided they treat the process with the same seriousness as a traditional IPO and maintain proper quarterly reporting and governance standards. 3. Enterprise customers are developing sophisticated strategies to reduce dependency on expensive frontier AI models, fundamentally challenging Anthropic's growth projections. AT&T serves as a prime example, having implemented an internal routing system using Light LLM that standardizes most programming tasks to cheaper Chinese open-source models while reserving Anthropic's Claude only for the most complex edge cases. This approach has resulted in flat growth from AT&T rather than continued expansion, suggesting that as enterprises mature in their AI usage, they will increasingly commoditize routine tasks and limit expensive API usage. This pattern mirrors the gradual enterprise adoption of SaaS between 2010 and present, where initial resistance gave way to widespread adoption, but ultimately led to sophisticated cost management strategies. 4. There is significant cultural disconnect between Anthropic's leadership and its user base that could undermine long-term customer loyalty. Anthropic operates as what could be described as a cult-like organization centered around machine learning engineers who subscribe to rationalist philosophy and beliefs in technological singularity, stemming from Eliezer Yudkowsky's writings including "Harry Potter and the Methods of Rationality." This creates a situation where the company treats non-machine-learning-engineer users, including paying subscribers and enterprise customers, as essentially irrelevant to their ultimate mission. Examples like their confusing communication about rate limit increases demonstrate this disconnect, fostering ill will among users who feel treated like children rather than valued customers, though this may not matter as much for enterprise contracts managed by sophisticated IT departments. 5. The rise of local LLM inference on consumer hardware threatens to disrupt the centralized API model that underpins Anthropic's revenue projections. With Apple M-series Mac Minis now available for lease at 200 dollars per month, the same price as Anthropic's premium subscription, and capable of running powerful open-source models with only electricity costs, a fundamental shift is occurring in the client-server dynamic of AI services. Apple is positioning itself as the dominant player on the client side with on-device AI capabilities, while Anthropic focuses on server-side inference, but as more developers and power users gain the technical sophistication to run local models, the question becomes whether Anthropic can maintain its premium pricing and growth trajectory when an increasing portion of inference moves to edge devices. 6. Dual-class share structures have become the standard governance model for founder-led technology companies going public, prioritizing visionary control over traditional shareholder governance. Following examples set by Google, Facebook, Palantir, and SpaceX, Anthropic is expected to implement a dual-class structure where Dario Amodei and potentially his sister retain voting control despite owning a minority of economic interest. This structure, where Class B shares carry 10 to 20 votes each compared to one vote for publicly traded Class A shares, allows founders to pursue long-term visions without quarterly pressure from institutional shareholders. While this prevents the collective shareholder activism that historically provided governance oversight, it has proven successful for iconic CEOs like Mark Zuckerberg, though it fundamentally changes the relationship between public companies and their investors compared to the traditional structure from the 1980s and 1990s. 7. The current AI boom represents an acknowledged bubble, but one without clear historical precedent for predicting its resolution or timeline. The traditional rule that "if everyone is asking whether we're in a bubble, we are, and if everyone is trying to predict when it ends, it already has" is being violated because people have been calling this a bubble for three years without resolution. The desperation of both Anthropic and OpenAI to go public quickly reflects internal recognition that the window may close, with both companies racing to raise massive amounts of public capital before market conditions change. Unlike the 2001 dot-com crash, the current situation lacks clear precedent because the technology is genuinely transformative and revenue growth is real rather than purely speculative, making it impossible to predict when or how the bubble might deflate, though the concentration of capital in a few mega-IPOs suggests the market dynamics have fundamentally changed from previous technology cycles.

  • S1 · E105
    September 3 · 1 hr 1 min

    Episode #105: Beating OpenAI at Their Own Game: Our Case, and the Filing That Settles It

    In this episode of the Stewart Squared podcast, host Stewart Alsop III and his father Stewart Alsop II dig into the unprecedented AI infrastructure buildout happening across big tech, with Google raising $85 billion in equity and floating bonds to fund data centers while questions mount about whether this mirrors the dark fiber overbuilding of 2001. They analyze the diverging paths of Anthropic and OpenAI, with Anthropic's revenue reportedly hitting a $65 billion run rate—now surpassing OpenAI—while maintaining a focused enterprise strategy versus OpenAI's scattershot approach and slowing growth. The conversation covers everything from token economics and the commodification of frontier models to SpaceX's Starlink business carrying the company, OpenRouter's role as the infrastructure layer enabling efficient model switching (recently acquired by Stripe for $7.5 billion), and predictions around the upcoming Anthropic IPO expected in September, which could beat OpenAI to market and fundamentally shift the competitive landscape. Timestamps 00:00 Discussion begins on AI foundational models' unprecedented fundraising for data centers, with Google raising $80 billion in equity and floating bonds to build infrastructure. 05:00 Examining Berkshire Hathaway's $10 billion investment in Google and comparing data center strategies between Anthropic's measured approach versus OpenAI's aggressive buildout plans. 10:00 Exploring token consumption efficiency and the mythical man month concept as LLMs begin programming themselves, creating uncertainty around future productivity metrics. 15:00 Analysis of Stripe acquiring OpenRouter for $7.5 billion and the virtualization of AI infrastructure enabling users to switch between multiple models seamlessly. 20:00 Discussion of OpenRouter's functionality as a platform for routing between models and how Anthropic attempted but failed to shut down programmatic API access. 25:00 Examining token pricing strategies as both Anthropic and OpenAI reduce promotional discounts while Grok introduces competing terminal products to maintain user costs. 30:00 Comparing streaming service fragmentation to LLM competition and the opportunity for intermediary services that prioritize users over platforms in enterprise deployment. 35:00 Analysis of United Airlines' real-time IT transformation demonstrating how enterprise technology deployment creates competitive advantages through better customer service. 40:00 Breaking down Anthropic's revenue model combining per-seat subscriptions, per-token API usage, and dedicated capacity sales targeting enterprise customers effectively. 45:00 Discussing Anthropic's accelerated IPO timeline for September versus OpenAI's delayed October offering as revenue growth and strategic clarity create market differentiation. 50:00 Explaining S-1 filing process and roadshow mechanics for IPOs while comparing SpaceX's successful trillion-dollar offering to anticipated Anthropic public debut. 55:00 Exploring SpaceX's Starlink economics and satellite business growth driving profitability while government contracts triple under current administration supporting rocket operations. Key Insights 1. Major tech companies are conducting unprecedented fundraising to build AI data centers, with Google raising $80 billion in equity for the first time since 2005 and also floating bonds. This massive capital expenditure mirrors the early 2000s dark fiber buildout when telecom companies overbuilt infrastructure and subsequently went bankrupt, raising questions about whether AI data centers will face a similar fate. The scale of spending is completely unprecedented, with companies spending beyond their substantial cash flows to build out capacity, making data centers an increasingly contentious political issue as communities resist their presence due to noise, water usage, and power consumption concerns. 2. Chinese AI investment is growing faster than American spending despite starting from a smaller base. Goldman Sachs projects that Alibaba, Tencent, ByteDance, and Baidu will spend roughly $102 billion combined in 2026, growing at over 80 percent year over year, while the American big four companies are spending approximately $764 billion with 77 percent growth. Although the United States currently maintains a significant spending advantage, China's accelerating investment rate and the overall demand dynamics create complex forecasting challenges similar to the fiber buildout era, where excess capacity took ten to fifteen years to be fully utilized. 3. Anthropic has emerged as the clear winner in the foundational model race, with revenue hitting a $65 billion annual run rate by July 2025, up sevenfold from the previous year. The company's focused enterprise strategy, selling primarily to businesses rather than consumers, contrasts sharply with OpenAI's scattered approach. Anthropic maintains three revenue streams: per seat pricing for application products, per token pricing on APIs, and dedicated capacity for large steady workloads. This disciplined approach to building infrastructure that matches revenue growth, rather than OpenAI's strategy of building maximum capacity and hoping demand catches up, positions Anthropic favorably for a potential $100 billion IPO as early as September. 4. OpenAI faces significant strategic challenges that have slowed its growth and delayed its IPO plans from the originally anticipated October timeframe to an uncertain future date. The company lacks a coherent product strategy, bouncing between initiatives while its revenue growth has decelerated compared to Anthropic. The organizational structure appears fragmented, with CEO Sam Altman, CFO Sarah Fryer, and President Greg Brockman not operating as a unified leadership team. If Anthropic successfully completes its IPO first, it will set unfavorable comparison terms for OpenAI, potentially creating an unmitigated disaster scenario where investors question why they would buy OpenAI stock when Anthropic demonstrates superior execution and faster growth. 5. Enterprise adoption of AI represents a fundamental shift comparable to past technology transitions, with companies like United Airlines demonstrating how real time AI powered systems create competitive advantages. The mythical man month principle that governed programming efficiency for decades no longer applies when AI agents can write code with minimal human supervision. However, token consumption and cost management have become critical issues, with companies like ATT moving 60 to 70 percent of their AI workloads from premium models to open source alternatives to control expenses. Both Anthropic and OpenAI recently ended promotional discounts, signaling a shift from market share acquisition to profitability, similar to how Uber eventually raised prices after establishing market dominance. 6. The AI infrastructure ecosystem has become highly virtualized and complex, with services like OpenRouter enabling seamless switching between multiple models and providers. This virtualization creates opportunities for intermediary services that manage costs and optimize model selection, similar to how streaming services compete for viewer attention. However, model providers like Anthropic attempted to shut down programmatic API access when services like Claude Bot emerged, only to reverse course a month later when they recognized their models had become commoditized. The fundamental challenge is that frontier models are rapidly becoming commodities as Chinese competitors and open source alternatives close the capability gap. 7. SpaceX successfully executed an unprecedented IPO exceeding one trillion dollars in valuation by raising $75 billion, setting new precedents for mega offerings that will influence how Anthropic and OpenAI approach their public market debuts. The SpaceX offering demonstrated that companies can maintain high valuations through diversified revenue streams, with Starlink satellite internet service proving highly profitable and growing rapidly while subsidizing other business segments. The company strategically rents excess data center capacity to competitors like Anthropic, generating unexpected revenue that supported its valuation. This model suggests that AI companies may need similar diversification strategies rather than relying solely on foundational model development to sustain public market valuations.

  • S1 · E104
    August 27 · 1 hr 8 min

    Episode #104: 36,000 Companies, One Metric: What DPI Did to Private Equity

    In this episode of the Stewart Squared podcast, host Stewart Alsop sits down with his father Stewart Alsop II to unpack how private equity, venture capital, and growth equity have evolved—and possibly converged—since the 2008 financial crisis. They explore how massive liquidity injections in 2020 fueled a PE buying spree, why the distinctions between investment categories are blurring as everyone gets measured by the same metrics (DPI—distributions per investment), and whether the whole system is starting to break down. The conversation touches on everything from SPACs making a comeback to Elon Musk's SpaceX IPO, the financialization of restaurants, and how even the meaning of terms like "bank" and "cash" are shifting in real time. In the final segment, they bring on surprise guest Tommy Yu, CEO and founder of TurnOn Technologies, to discuss closed-loop payment systems, stored value, and why traditional finance people struggle to see beyond Visa and Mastercard logos even when companies like Starbucks are sitting on $2 billion in unredeemed gift card float. Timestamps 00:00 Introduction and experimental format with a surprise guest, discussing private equity shifts from 2008 through 2023 and massive liquidity changes fueling PE buyouts 05:00 Historical perspective on venture capital evolution starting from the seventies, pension fund rule changes allowing risky investments, and the emergence of hedge funds and private equity differentiation 10:00 Growth equity versus private equity distinctions, SPACs making a comeback after previous failures, and the shrinking number of public companies from 7000 to 4000 15:00 The fundamentals of capitalism and time value of money, government money printing increasing liquidity, and how pension funds now invest heavily in alternative assets 20:00 DPI measurement becoming universal across all investment types, distinguishing between realized and unrealized gains, and how capital calling works in venture funds 25:00 Limited partnerships structure in venture capital, restaurant investments as different from tech startups, and Main Street versus Silicon Valley business models 30:00 Pension funds controlling massive percentages of national assets, asset allocation strategies across different investment vehicles, and everything being measured the same way now 35:00 How institutional definitions are breaking down, terms losing their original meanings, banks becoming something entirely different, and companies essentially functioning as banks themselves 40:00 Chinese centralized system appearing more effective than Western capitalism currently, enlightened dictatorship efficiency, and how professional attention demands are increasing dramatically 45:00 Vibe coding distinctions and keeping up with changing terminology, inviting guest Tommy Yu to discuss his fintech company TurnOn Technologies and stored value systems 50:00 Liberty Mutual's restaurant investment treating food as tradable assets, closed loop versus open loop systems like Starbucks' 8 billion dollar gift card float earning interest 55:00 Regulation falling behind fintech innovation, crypto remaining largely unregulated, and how intelligent investors can differentiate when everyone's doing the same thing 60:00 TurnOn Technologies creating universal stored value systems beyond single companies, neobanks as digital tellers, and younger generations caring more about user experience than traditional banking Key Insights 1. The venture capital and private equity landscape has fundamentally transformed since the seventies when pension funds were first allowed to invest in risky alternative assets. What began as distinct categories with different purposes and metrics has now blurred together, with all investment types being measured by the same standard called DPI or distributions per investment dollar. This measures how quickly investors get their capital back in cash, and the problem is that venture capital now takes fifteen plus years to return capital compared to the historical five to ten years, making it less attractive when banks offer five percent and public markets offer ten percent returns. 2. Private equity has exploded to encompass over thirty six thousand companies in the United States, far exceeding the roughly four thousand public companies that exist today, down from seven thousand in nineteen ninety six. This represents a massive shift where the ownership layer has moved increasingly into private hands, with companies staying private longer and accessing capital through growth equity and private equity rather than going public. The lines between venture capital, growth equity, and private equity have become so blurred that even experienced investors struggle to articulate meaningful differences between these categories. 3. The financial system is becoming increasingly opaque and unregulated despite the perception that finance is highly regulated. New categories like private credit have emerged as what some consider cesspools of activity that regulators cannot keep pace with or even understand. The regulatory framework has fallen so far behind the actual innovations in fintech, crypto, and alternative investments that there is effectively no meaningful oversight in many areas, creating opportunities for both innovation and potential abuse that would have been impossible in previous eras. 4. The fundamental terms and definitions that underpin capitalism are changing so rapidly that experienced investors and business people can no longer rely on historical understanding. What constitutes a bank, what money actually means, what liquidity represents, and even what a company is have all shifted dramatically. Cash itself has become metaphorical rather than physical, and businesses are increasingly functioning as their own banks by holding stored value and earning interest on customer deposits, as demonstrated by Starbucks running eight billion dollars through gift cards with two billion in unredeemed float. 5. The Chinese communist system under Xi Jinping is currently working better for its people than the capitalist system is working for citizens of capitalist countries, creating an existential challenge to the assumption that capitalism is the superior economic model. This represents a historic shift where an enlightened dictatorship with strategic central planning is outperforming the chaotic and unpredictable leadership in capitalist democracies. The comparison suggests that the ideological certainty about capitalism's superiority may need to be reconsidered in light of actual results for ordinary citizens. 6. The democratization of investing through new structures like SPACs and the accessibility of markets has created a situation where the distinction between legitimate investment vehicles and pyramid schemes has become uncomfortably narrow. The entire system increasingly relies on later investors buying out earlier investors at higher valuations, with the public markets serving as the final exit for private investors to realize gains. This raises uncomfortable questions about whether the fundamental structure of modern capitalism differs meaningfully from the Bernie Madoff scheme that collapsed fifteen years ago. 7. The restaurant industry and Main Street businesses represent a parallel economy that operates on completely different principles than Silicon Valley startups, yet investment vehicles are increasingly trying to treat them as comparable assets. Liberty Mutual investing three hundred twenty million dollars in restaurant food credits represents the financialization of everything, where even meals become tradable assets divorced from the underlying business reality. This demonstrates how the investment world is desperately seeking returns in increasingly exotic and risky categories as traditional distinctions break down an...

  • S1 · E103
    August 20 · 1 hr 1 min

    Episode #103: Scarce to Itself: NVIDIA, Apple, a Driverless Zoox, and the Real Fight Over the Future of Cars

    In this episode of Stewart Squared, Stewart Alsop III and Stewart Alsop II dig into the fast-moving world of self-driving cars — from Cruise's rocky history and Waymo's expansion to Zoox's driverless design and NVIDIA's growing role as the go-to OEM partner for automakers building software- and AI-defined vehicles — before the conversation branches into dual-use tech lessons from Ukraine, the US-China race in autonomy and semiconductors, the DRAM memory shortage squeezing Apple and the gaming industry, Slate's stripped-down electric truck, and a closing riff on performance-based, subscription-driven podcasting. Timestamps 00:00 — Self-driving cars and the fallout from Cruise’s collapse; where AVs are operating now, plus the role of NVIDIA in autonomous vehicles. 00:05 — Software-defined cars, Level 4 autonomy, and why Waymo and Zoox matter; how regulation and city-by-city permits shape rollout. 00:10 — Dual-use autonomy in war zones, liability, and the shift from consumer adoption to fleet adoption; NVIDIA’s place as a supplier, not a prime. 00:15 — Why NVIDIA is strategically positioned across chips, software, and cars; comparisons with Tesla, Rivian, Ford, and GM. 00:20 — China’s open-source AI push, RISC-V, control vs openness, and the tension between state control and innovation. 00:25 — The rise of neo-primes like Anduril, how the Pentagon buys systems, and why the U. S. defense market favors trust and scale. 00:30 — A broader debate on capitalism vs communism, China’s economic strain, and whether the future looks more centralized or decentralized. 00:35 — The memory crisis: DRAM, GPUs, Apple, and video games getting squeezed as AI demand drives prices up. 00:40 — Cheap EV disruption with Slate, new form factors, and how car design may split between utility-first and premium autonomous vehicles. 00:45 — The city itself changing: AVs, urban planning, flying-car ideas, and how autonomy could reshape Los Angeles and beyond. 00:50 — Pricing scarce compute, NVIDIA’s leverage, and the idea that some companies become scarce to themselves. 00:55 — A long-range view: centralized vs decentralized tech cycles, RISC-V, and the possibility of open hardware reshaping everything. Key Insights Autonomous vehicles are consolidating around a small set of platform players. Waymo, Tesla, Zoox, and Rivian lead the US market, while China's fleet is dominated by companies like Geely-backed operations. Regulation moves city by city, so scale depends as much on winning local approval as on the technology itself. NVIDIA has built a stealth position as the OEM backbone of autonomous vehicles. Rather than compete with carmakers, NVIDIA supplies the compute platform nearly every manufacturer relies on, letting it profit from the industry's growth without taking on liability or betting on any single winner. Liability, not technology, is the biggest brake on driverless fleets. Once a car has no human driver, responsibility shifts entirely to the fleet owner. Zoox's steering-wheel-free design in San Francisco is the clearest test case for how that liability question gets resolved. The DRAM shortage is a downstream effect of the AI buildout. Massive GPU demand for LLM training soaked up memory supply, driving up DRAM prices for everyone else — squeezing Apple, gaming consoles, and now the auto industry, which needs growing amounts of compute per vehicle. China's dominance in EVs coexists with deep structural weakness. A shrinking, aging population, price wars that erase margins in sectors like solar, and a government that reins in entrepreneurs when they get too powerful (DeepSeek's blocked IPO) complicate the narrative of unstoppable Chinese industrial growth. "Scarce to itself" describes a new kind of market power. Companies like NVIDIA and Apple aren't scarce because of demand tricks — they're supply-constrained on their own hardware, which lets them set pricing on their own terms rather than compete on it. Podcasting may be shifting from ads to direct subscription models. Inspired by ideas like Trump's paid early-access posts, Stewart Alsop and Stewart Alsop II are testing "performance podcasting" — charging listeners directly for real-time access instead of relying on sponsorships.

  • S1 · E102
    August 13 · 51 min

    Episode #102: AI Is Eating the World’s Memory

    In this episode of Stewart Squared, Stewart Alsop III and Stewart Alsop II sit into the shifting ground beneath the AI boom, starting with the strange saga of Leopold Aschenbrenner's hedge fund and the memory chip shortage that's rippling through everything from Apple's product line to the gaming industry, before moving into how OpenAI and Sam Altman's data center spending is reshaping global compute demand, the widening gap between American and international tech ecosystems, China's uneasy relationship with open source AI, and a look at Mira Murati's new venture Thinking Machines Lab and its fine-tuning tool Tinker, wrapping up with some thoughts on what a live, interactive version of the show could look like. Timestamps 00:00 AI bubble and the Leopold Aschenbrunner hedge-fund story; debt, margin pressure, and why memory stocks surged 00:05 memory becomes the bottleneck as AI data centers expand; Apple, Micron, and rising RAM prices 00:10 Korea takes a hit from memory-market swings; contrast with Japan, manufacturing, and the karetsu model 00:15 Live translation tech, Google’s new API, and a side discussion of robotics and Japanese PC history 00:20 Japan’s early PC ecosystem, NTT, Microsoft, Windows, and why standards won the market 00:25 Back to finance: equity vs debt, leverage, Glass-Steagall, and how banking got reorganized 00:30 AI hedge funds, risk, Citadel, margin calls, and the distinction between lending and investing 00:35 Capitalism by starting companies vs buying companies; why money is partly a metaphor 00:40 AI agents, Chrome, and the difference between distributed and federated systems 00:45 Matrix and Nostr, open-source messaging, and the internet’s new borders 00:50 Live audience questions, interactive publishing, and the business potential of real-time conversation Key Insights The AI boom is straining a memory chip supply that can't scale fast enough. Massive spending on AI data centers by companies like OpenAI has driven demand for DRAM through the roof, and because memory factories take years to build, prices have roughly tripled in six months — squeezing everyone from Apple (which is struggling to ship products like the Mac mini) to the broader computer gaming industry. Leverage is what turned a smart bet into a crisis. Leopold Aschenbrenner's "Situational Awareness" hedge fund quadrupled investor money early on by going heavy into AI, but borrowing tens of billions against volatile chip and memory stocks left him exposed when prices tanked — prime brokers like Bank of America, Goldman Sachs, and JPMorgan Chase issued margin calls, and Citadel ultimately bought the distressed assets at a steep discount. Korea's economy is deeply entangled with the memory business. Companies like SK Hynix built Korea into a manufacturing powerhouse for chips, and that same concentration made its stock market especially vulnerable — the market reportedly fell more than 33% in July, a decline worse than the crashes of 1997 and 2015. Deregulation reshaped modern finance in ways still being felt today. The conversation traces a line from Glass-Steagall's separation of commercial and investment banking, through its effective rollback via the Gramm-Leach-Bliley Act, to today's financial holding companies (like JPMorgan owning both Chase and an investment bank) — blurring risk in ways that echo the AI/memory borrowing spiral. National tech ecosystems don't automatically follow American patterns. Japan's PC industry, dominated by NTT, never fully converged on the IBM-clone standard the U.S. market did, and Korea's cultural relationship to gaming and digital life (partly shaped by heavy state investment in nationwide internet infrastructure) diverged sharply from its neighbors, despite a shared heritage. China's relationship with open-source AI is shifting. Having initially embraced open source partly because it seemed easier to control than proprietary Western tech, China now appears increasingly concerned about having lost that control and is working to reassert it. Distributed, interoperable messaging protocols are having a moment. Tools like Matrix and Nostr (open-source alternatives gaining traction partly in response to Meta's restrictions) revive a decades-old dream of software systems messaging each other freely, distinguishing distributed "pub-sub" models from federated ones where all participants must cooperate.

  • S1 · E101
    August 6 · 54 min

    Episode #101: Apple's AI Is Finally Here. Why Does It Still Feel Broken?

    In this episode of the Stewart Squared podcast, hosts Stewart Alsop and Stewart Alsop II dig into Apple's rocky iOS 27 rollout and the Apple Intelligence features that still don't quite work, before spiraling into Apple's org chart and headcount, the lost art of building apps, chip design and the Apple 100, open source versus closed source (MLX, the Linux kernel, GitHub vs. GitLab), Claude Code and why Anthropic's terminal-first approach might be winning the AI race, the commoditization debate around Chinese open source models, Adobe's fall from grace and the Postscript-to-Flash saga with Steve Jobs, real-time publishing and the Ben Thompson model for podcasting, and manufacturing hardware from PCBs to Sonos speakers. Timestamps 00:00 — iOS 27 rollout and buggy Apple Intelligence features frustrate both hosts. 05:00 — Debating Apple's headcount, retail vs. corporate split, and designers fleeing to OpenAI. 10:00 — The Apple 100, the blurry line between research and development, and Xerox PARC. 15:00 — Apple's custom chips, open source roots like the Linux kernel and MLX. 20:00 — GitHub origins, Microsoft's enterprise mentality, and life since MS-DOS. 25:00 — Claude Code, Boris Cherny, and why terminal agents are reshaping coding. 30:00 — AI's text-based limits, Chinese open source models, and a coming robotics interview in Japan. 35:00 — GitLab vs. GitHub, what a workbench and compiling actually mean, and Mac performance gripes. 40:00 — Postscript vs. TypeScript, the Courier font, and early PageMaker newsletters. 45:00 — Hot type, the printing press, and why neither host went into industrial robotics. 50:00 — Editing Marine Business magazine and watching Japan and China take over manufacturing. 55:00 — Building PCBs, lessons from Sonos, and pitching a pay-to-listen real-time model. Key Insights Apple's biggest weakness isn't hardware or chips—it's software. Despite two years of hype, Apple Intelligence still creates duplicate calendar events and can't recognize things already scheduled, revealing a company that excels at silicon and operating systems but consistently ships mediocre apps, a gap the hosts trace back to Tim Cook's leadership. Apple's culture runs on a quiet meritocracy called the "Apple 100," a Steve Jobs-era concept where influence isn't tied to title—a junior engineer can be as pivotal as an executive, which explains how the company sustains innovation despite a bloated headcount of roughly 166,000, nearly half of it in retail. Anthropic's edge may not be model quality alone but its decision to build Claude Code around the terminal, treating programming as just another form of text prediction. This bet, credited largely to Boris Cherny, let Anthropic reach developers directly rather than waiting for polished consumer products. The commoditization narrative around AI cuts both ways. As Chinese open-source models close the gap with American closed-source ones, it either means nobody can maintain a lasting lead, or—as one host argues—the opposite: that leaders become nearly impossible to catch once compounding advantages set in. Adobe's arc from a lean systems company to what one host calls a fallen giant shows what happens when a company loses its performance-driven roots. Built on Postscript and page-description technology for the LaserWriter, Adobe eventually prioritized cross-platform reach over speed, echoing Apple's own struggles with app quality. Real-time publishing is emerging as a business model, not just a technical curiosity. Drawing on Ben Thompson's subscription-driven podcast network, the hosts float charging listeners for live access, turning the current ten-day publishing delay from a limitation into a monetizable feature. Manufacturing know-how doesn't transfer easily across domains. Lessons from Sonos scaling hardware in China, and earlier stories from Mercury Marine's engine factories, show that going from prototype to mass production—especially with physical components like PCBs and speakers—demands specialized expertise that even seasoned tech investors admit they lack.

  • S1 · E100
    July 30 · 1 hr 4 min

    Episode #100: From Apple's iOS 27 to Anduril's Defense Tech: Where AI's Advantage Really Lies

    In this episode of Stewart Squared, Stewart Alsop III sits down with his father and co-host Stewart Alsop II for a wide-ranging conversation that jumps from Apple's iOS 27 preview beta and the long road to Apple Intelligence, to the trust gap between Anthropic and OpenAI and the rise of digital-twin apps like Sentience, before pivoting into venture capital territory with a candid look at Andoril, Palmer Luckey, and the defense-tech boom reshaping how the primes do business; from there the two work through the surveillance creep of modern police tech, China's near-peer standing against the U.S. and its own reusable-rocket ambitions, and finally land on the state of fintech trust, the SpaceX IPO, and how thirty years of early-stage deal-making stack up against today's AI-driven venture landscape. Timestamps 00:05:00 — Apple Intelligence and the iOS 27 beta merge AI with hardware for an always-on assistant vision. 00:10:00 — Anthropic vs. OpenAI trust, Mira Murati's open-source push, and the Sentience digital-twin app. 00:15:00 — Apple's on-device security compared against Google and Microsoft. 00:20:00 — Tech billionaire philanthropy and legacy: Gates, Zuckerberg, and Jobs. 00:25:00 — Andoril and the personal story of investing alongside Palmer Luckey. 00:30:00 — History of defense-tech venture capital and Andoril's government ties. 00:35:00 — Palantir expanding into Argentina and the rise of predictive policing. 00:40:00 — Data, information, and wisdom in AI-driven knowledge management. 00:45:00 — Token minimizing strategy for running Claude and Codex coding agents. 00:50:00 — Fintech trust: Stripe, Venmo, PayPal, and the Panama Papers. 00:55:00 — SpaceX's IPO and public market valuation reflections. 01:00:00 — Reflexivity, Soros, and LLM token economics shaping VC decisions. Key Insights Apple's iOS 27 beta shows the company finally following through on the Apple Intelligence promise it botched two years ago, and its real advantage isn't the AI itself but that it's fused to hardware holding a user's calendar, messages, and contacts, letting it answer deeply personal questions Google can't match outside its own Pixel devices. Anthropic's positioning as "the Apple of AI" reflects a market increasingly sorting by trust rather than raw capability, with younger users drifting toward open-source alternatives like Mira Murati's newly funded startup, suggesting safety-focused branding alone won't hold loyalty across generations. Apps like Sentience, which build a "digital twin" by ingesting years of email, messages, and calendar history, hint at where personal AI is headed, but the gap between their mobile and desktop functionality shows this category is still early and unevenly built. Venture capital has quietly become the primary funder of military innovation, with firms like Founders Fund turning early bets on companies such as Andoril and Palantir into a broader industry rush toward defense and dual-use technology after decades of stagnant, cost-plus contracting among the traditional prime contractors. Predictive policing tools are reinforcing existing patterns rather than improving outcomes, since they're trained on historical data that sends more patrols into already over-policed neighborhoods, raising questions about transparency as governments adopt surveillance faster than citizens can question it. The venture capital game has shifted dramatically from early-stage bets to massive growth-equity checks, with average valuations jumping roughly ninetyfold over a decade as AI and defense deals now routinely reach into the billions, leaving classic early-stage investors feeling sidelined by their own industry. As AI agents multiply, the real competitive edge is shifting from model performance to token efficiency, with a "token minimizing" approach using multiple coding agents in parallel emerging as a practical way to build software at scale without hitting rate limits or runaway costs.

  • S1 · E99
    July 23 · 37 min

    Episode #99: Can Money Buy Meta a Comeback in AI?

    In this episode of the Stewart Squared podcast, host Stewart Alsop and guest Stewart Alsop II dive into Meta's latest AI model releases and their broader issues with user addiction, touching on the European Commission's warnings about addictive features and massive fines totaling $1.4 trillion from US state attorney generals. The conversation ranges from Meta's Meta Super Intelligence Lab and their attempts to catch up to OpenAI and Anthropic, to the impossibility of governments controlling AI development as countries rush to build sovereign models. They discuss NVIDIA's open source robotics models, debate the future of humanoid versus non-humanoid robots, and compare the business approaches of Mark Zuckerberg and Elon Musk. The episode also covers Trump's floating ideas about restricting state-of-the-art AI models to US citizens, China's similar restrictions, SpaceX's recent IPO performance, and the concept of shareholder capitalism as applied to government investments in tech companies like Intel and potentially OpenAI. Timestamps 00:00 Meta releases new AI model and thought-reading technology while facing trillion-dollar fines from state attorneys general for social media harm, particularly to young people 05:00 Discussion of Meta's superintelligence lab attempting to catch up with OpenAI and Anthropic, while their cash-harvesting social media business funds AI development despite past VR failures 10:00 Government inability to regulate fast-moving AI technology, with Trump and China floating ideas about restricting state-of-the-art models to citizens only 15:00 Examining how AI's addictive nature combined with existential fears creates political volatility, plus NVIDIA's open-source robotics models becoming viable alternatives 20:00 Debate over humanoid versus non-humanoid robots, discussing industrial applications and questioning whether humanoid design makes practical sense for factories or homes 25:00 Comparing Elon Musk's technical accomplishments at Tesla and SpaceX with Zuckerberg's social media empire, noting Facebook's real-time scaling innovation happened decades ago 30:00 SpaceX AI IPO analysis predicting failure if stock drops below offering price, plus discussion of shareholder capitalism and Trump's government investment strategy 35:00 Reflecting on information overload in the AI age making it impossible to understand complexity, with neither Trump nor technological developments being predictable anymore Key Insights 1. Meta faces massive legal liability for its addictive social media practices, with state attorney generals demanding approximately 1.4 trillion dollars in total penalties, including a New Mexico jury awarding 375 million dollars in civil penalties and the state separately seeking 2.7 billion dollars in abatement costs. The European Commission has warned Meta about continued use of addictive features, though any meaningful fine would need to be extraordinarily large given Meta's 2 trillion dollar valuation and substantial cash flow. Despite these legal challenges, Meta continues to harvest cash at an astonishing rate from Instagram, Facebook, and Threads by addicting users without regard for their wellbeing, using that revenue to fund their artificial intelligence initiatives after wasting money on virtual reality. 2. Meta's artificial intelligence efforts through their Meta Super Intelligence Lab have been mixed, with their initial LAMA 4 model considered a disaster, but some internal evaluations suggest their upcoming release could potentially help them catch up to OpenAI and Anthropic, possibly even displacing Google as the third accepted foundation model. The key difference between Meta and competitors like OpenAI and Anthropic is that Meta has enormous cash flow from their social media properties to support their AI development, allowing them to spend freely even if they waste money, whereas OpenAI and Anthropic only generate revenue from their AI products. However, there remains skepticism about whether Meta can truly catch up once having fallen behind in the competitive landscape of artificial intelligence development. 3. Governments are fundamentally irrelevant in controlling artificial intelligence development because technology moves too fast for governmental bodies to understand or regulate effectively. Both Trump and China have floated ideas about restricting state-of-the-art AI models to their respective citizens, but these efforts cannot succeed because AI models are infinitely copyable and open source models are becoming increasingly powerful. The reality is that Pandora's box is already open with AI technology, and as countries realize they don't want dependence on China or the United States, they will develop their own sovereign models, creating a mushroom effect that makes control impossible regardless of what governments attempt to mandate or regulate. 4. NVIDIA is becoming increasingly important in the open source AI model space, particularly for robotics applications, as they develop small open source models that can run inside robots without requiring NVIDIA to monetize the models directly since they profit from hardware sales. Jensen Huang has publicly stated that robotics represents the next major innovation, leading NVIDIA to focus on developing CPUs alongside GPUs and integrated systems with small models for physical AI applications. This represents a significant shift where developers no longer need to rely solely on Chinese models, as NVIDIA's open source offerings are becoming genuinely competitive and useful for specialized applications like machine learning cameras and embedded robotics systems. 5. The definition and future of robotics remains highly contested, with significant debate between those advocating for humanoid robots versus non-humanoid specialized robots, and the Wall Street Journal recently published analysis suggesting humanoid robots may not be the optimal path forward. Tesla has been successful partly because they integrated industrial robots from the beginning rather than hand-building cars, reducing production costs substantially, though their humanoid robot demonstrations have not yet resulted in actual factory deployment despite ambitious forecasts. The challenge with humanoid robots includes safety concerns like a hundred-pound robot potentially killing a child if it falls, and the complexity of replicating human capabilities like hands, though companies like Neo recently claimed to have developed hands that work better than humans. 6. The comparison between Mark Zuckerberg and Elon Musk reveals stark differences in technical accomplishment, with Zuckerberg's primary innovation being real-time scaling for billions of users achieved around 2007, after which Facebook has largely exploited that technology to extract money without meaningful additional innovation. In contrast, Elon Musk has accomplished multiple extraordinary technical achievements simultaneously including getting people to buy Teslas, building factories for cars and batteries, changing the distribution system to bypass dealers, building an electric charging network, and creating SpaceX and Starlink. While Musk may be crazy and hard to like, he has genuinely accomplished substantial technical innovations across multiple domains, whereas Zuckerberg has primarily focused on corrupting youth and harvesting data for the past sixteen years. 7. The SpaceX AI initial public offering illustrates important dynamics about public market trust and company valuation, with shares issued at 135 dollars now trading around 145 dollars after initially rising but falling back near the offering price, and predictions suggest it may fall below the offering price before lockup periods expire. The float representing publicly traded shares is only about five percent of total shares, and when more shares become available ...

  • S1 · E98
    July 16 · 57 min

    Episode #98: What Apple Gets That the Rest of Tech Doesn't: Trust Scales

    In this episode of the Stewart Squared podcast, host Stewart Alsop sits down with his father, guest Stewart Alsop II, to tackle a wide range of tech topics from AI chip design to cybersecurity vulnerabilities. The conversation covers OpenAI's Jalapeno chip (trained by AI in just nine months), the emerging etched.com platform, and Cloudflare's recent power move against Google, while Stewart shares his experience building real-time games on video calls and experimenting with ESP32 hardware for robotic projects. The discussion also dives into Meta's controversial KYC (know your customer) requirements that got Stewart kicked off Facebook and Instagram, Apple's evolution from hardware company to trusted computing partner under Tim Cook's leadership, the security implications of Chinese-manufactured ESP32 chips, and why hardware-focused companies struggle to adopt AI-driven development practices like using LLMs to eliminate software bugs—all wrapped up with an AI fact-check of their previous episode's claims. Timestamps 00:00 Stewart welcomes listeners and mentions his father's return from travels while he's been enjoying winter in Buenos Aires, setting up discussion topics including etched.com, OpenAI's Jalapeno chip training, and Cloudflare's competitive moves against Google 05:00 Discussion shifts to Meta's controversial KYC implementation and Supreme Court decisions allowing Facebook to require identity verification, with Stewart expressing strong opposition to Meta's practices and considering abandoning their platforms except WhatsApp 10:00 Conversation explores Mark Zuckerberg's personality and Meta's toxic culture, comparing their approach to Apple's user-protective stance and examining how tech companies handle personal data and privacy differently across their platforms 15:00 Deep dive into Apple's historical positioning as user-friendly company under Steve Jobs and Tim Cook, discussing their discipline in product management and how they've maintained consumer trust through consistent privacy protection over decades 20:00 Exploration of hardware complexity and software challenges, including Stewart's robotics workshop using ESP32 microcontrollers where even experienced engineers struggled with basic connectivity issues highlighting system complexity 25:00 Analysis of why hardware companies struggle adopting AI solutions, discussing Apple's bug management approach and questioning why they don't leverage tools like Anthropic's Mythos for eliminating persistent software bugs systematically 30:00 Security architecture discussion focusing on Apple's Unix-based kernel foundation inherited from NeXT, explaining how Avie Tevanian built security into macOS from the beginning making Apple relatively breach-free compared to competitors 35:00 Linux and Unix history explored, examining open source security models and discussing ESP32 operating systems, revealing that FreeRTOS provides embedded operating system functionality for these Chinese-manufactured development boards 40:00 Chinese semiconductor company Espressif discussion, examining potential vulnerabilities in using Chinese hardware while distinguishing between chip-level security and application-layer data access risks in connected devices 45:00 Device Authority company case study about remote device validation and firmware updates, connecting to historical cyberattacks like Stuxnet virus that physically infected Iranian centrifuges without internet connectivity 50:00 Cybersecurity industry overview mentioning Israeli company Check Point as pioneering firm, emphasizing importance of hiring specialized security experts rather than attempting DIY cybersecurity for critical business applications 55:00 Fact-checking segment reviewing previous episode claims about Dario Amodei's credentials, NVIDIA founding dates, software patents, OpenAI's Jalapeno chip timeline, and Waymo's highway incidents with minor corrections noted throughout discussion Key Insights 1. Apple has maintained user trust through a fundamental alignment with individual privacy rather than corporate interests. Unlike companies such as Meta and Microsoft, Apple has built its brand on protecting user data and maintaining security at the operating system level. This cultural commitment, formalized under Tim Cook but rooted in Steve Jobs' vision, has given Apple a distinct competitive advantage with over 2.5 billion users who feel the company is genuinely on their side rather than exploiting them for advertising revenue or data harvesting. 2. Meta is conducting controversial Know Your Customer verification processes that may represent a troubling expansion of identification requirements for social media platforms. Following what appears to be a 2025 Supreme Court decision, Facebook and Instagram are implementing KYC protocols previously reserved for financial institutions, potentially to legally collect identifying information for AI model training. This practice has driven some users to abandon Meta platforms entirely, viewing it as an unacceptable intrusion that violates the original spirit of personal computing. 3. Hardware-focused companies struggle to adopt AI coding tools because their engineering culture emphasizes control and deterministic systems. Companies like Apple, despite their technical sophistication, remain slow to implement AI solutions for tasks like bug elimination because their hardware-oriented workforce consists of control-oriented engineers uncomfortable with the probabilistic nature of AI systems. This cultural resistance prevents them from fully leveraging tools that could theoretically eliminate persistent software bugs that have plagued their ecosystem for years. 4. Security architecture fundamentally differs between operating systems, with Unix-based systems maintaining inherent advantages. Apple's security strength derives from the Unix kernel inherited from NeXT in 1997, which was designed with security as a core principle. This foundation underlies all Apple operating systems today, from macOS to iOS. In contrast, Microsoft's Windows has never achieved comparable security, making it constantly vulnerable to exploitation despite being the standard for government systems, which represents a significant ongoing risk. 5. The Chinese technology ecosystem, particularly in embedded systems and semiconductors, presents complex security considerations that are more political than technical. Companies like Espressif, which manufactures the ESP32 microcontroller chips, are headquartered in Shanghai and dominate the affordable IoT device market. However, because much of this technology uses open source software like Linux, the actual security risks are less about the hardware itself and more about higher-level software implementations that could potentially access personal data, making concerns somewhat overblown outside of China's Great Firewall. 6. The transition from traditional software development to AI-assisted coding is democratizing hardware prototyping in unprecedented ways. Non-engineers can now successfully program microcontrollers and build functional robotic systems using AI coding assistants, while ironically, experienced electrical and software engineers sometimes struggle with the same tasks due to their ingrained approaches. This represents a fundamental shift in who can participate in hardware development, though it also introduces new considerations around security and trustworthiness of the resulting systems. 7. Modern cybersecurity remains a constant race between attackers and defenders who possess equivalent knowledge and capabilities. The distinction between white hat and black hat hackers is merely one of intention rather than skill, as both groups operate with the same information simultaneously. Historical examples like the Stuxnet attack on Iranian centrifuges demonstrate tha...

  • S1 · E97
    July 9 · 1 hr 3 min

    Episode #97: How AI Is Rewriting the Rules of Computing

    Stewart Alsop sits down with his father, Stewart Alsop II, to unpack what chip design even means anymore, starting with OpenAI's Jalapeno chip and the wild claim that it was designed in nine months using their own LLM. They trace the CPU from its personal computer origins through GPUs, FPGAs, and the strange new world where anyone might vibe code a chip, then swing into digital projection and showrunner systems at TeamLab and Meow Wolf, autonomous vehicles and the LiDAR fight between Tesla and Waymo, Gaussian splats and world models, a detour into 3D printing and a failed IRL collectibles startup, and a closing stretch on patents, IP trolls, and whether China's open source AI push means the US proprietary model is losing ground. Timestamps 00:00 Apple chips, CPU vs GPU and why chip design feels virtualized now. 05:00 GPUs for video games, why productivity ignored graphics, and how the CPU became a bundle of multiple cores. 10:00 Team Lab and Meow Wolf as digital-projection worlds: projectors, microcontrollers, and the showrunner idea. 15:00 Interactive exhibits as “everything at once,” then a shift into Anthropic, biotech, and the pace of AI innovation. 20:00 Real-time systems, lipsync, world models, and why LLMs struggle with space-time. 25:00 Autonomous vehicles: Cruise, Waymo, LIDAR, Tesla’s camera-only approach, and the debate over edge cases. 30:00 More on Waymo vs Tesla, safety incidents, and whether Gaussian splats matter for robotics. 35:00 Chip design, OpenAI’s “jalapeño” chip, firmware, memory shortages, and why Apple memory costs are rising. 40:00 Patents, software IP, LLMs, and how China and the US diverge on open source versus proprietary AI. Key Insights The CPU has quietly become plural. What used to be a single processing unit is now many cores managing memory, disk, and networking all at once — the concept of "central processing" has essentially been virtualized from the inside out. Chip design may no longer require deep technical expertise. OpenAI's Jalapeno chip, reportedly designed in nine months using their own LLM, suggests that designing silicon is becoming something closer to "vibe coding" than a specialized engineering discipline. Digital projection systems like TeamLab and Meow Wolf run on lightweight computing, not heavy processing power. The magic comes from networked microcontrollers and a "showrunner" system, a concept borrowed from television, that keeps hundreds of projected events in sync without conflict. Tesla and Waymo represent two opposing bets on autonomy. Tesla relies purely on cameras and processing power, while Waymo loads its cars with LiDAR, radar, and video. Both approaches still hit real-world edge cases, from Waymo pulling cars off freeways after construction-zone incidents to a fatal Tesla crash with no clear explanation. World models are trying to give machines a sense of space and time. Gaussian splats, used by companies like Marble and Niantic, create detailed spatial reconstructions, but they're not yet real-time, which limits how directly they can be applied to something like robotic driving. Intellectual property often only reveals its value after failure. A collectibles startup pairing physical figurines with digital twins collapsed alongside the NFT market, but the conversation underscores how "IP trolls" and specialists like Nathan Myhrvold later mine failed patents for value nobody recognized the first time around. China's AI progress is closing the gap through an open source strategy the US mostly abandoned. Coupled with Anthropic's accusation that Alibaba scraped its codebase millions of times, the episode frames China's non-profit-driven, open approach as a real competitive threat to America's proprietary model.

  • S1 · E96
    July 2 · 46 min

    Episode #96: From Steve Jobs to AI: The Stories That Never Became Data

    In this episode of Stewart Squared, Stewart Alsop III sits down with his father, Stewart Alsop II, for a wide-ranging conversation that moves from a heartfelt tribute to the late Brent Schlender — legendary tech journalist and author of Becoming Steve Jobs — through the history and philosophy of journalism, the concept of the fourth estate, and what it meant to cover Silicon Valley's biggest names up close. The two also dig into Cold War history, Russia's ambiguous relationship with the West, the Alsop family's own journalism legacy, and how AI is reshaping the way we think about memory, personal data, and the historical record. For more on Brent Schlender's work, check out his book Becoming Steve Jobs and Stewart Alsop II's Substack obituary for Brent, where he also shared the iconic Fortune magazine cover featuring Steve Jobs and Bill Gates together. Timestamps 0:00 Brent’s memorial and the jet lag opening, then into who Brent was and his role in tech journalism 5:00 Brent’s journalism background, friendship with major tech figures, and the idea of the three Steves in Steve Jobs’ story 10:00 Why journalists usually stay objective, what the fourth estate means, and how the press acts as a check on power 15:00 The press, patriotism, Cold War context, CIA tensions, and how journalists like Stuart and his uncle navigated American loyalty 20:00 McCarthyism, fear, false accusations, and how brave reporting protected people and challenged demagoguery 25:00 Russia as part West / East, Christianity, borders of identity, and the discussion shifting into Russian history 30:00 Soviet-era travel, tech speeches, old-school publishing, and the problem of reconstructing the past without a digital trail 35:00 History vs. journalism, archives, memory, and why preserving records matters for telling the story later 40:00 Brent’s memorial memories, the Steve Jobs book, and how Brent’s work shaped the industry through insight and relationships Key Insights Brent Schlender stood apart from most journalists because he became genuinely close friends with the people he covered — Steve Jobs, Bill Gates, Larry Ellison — and that access gave him a depth of understanding that produced what many consider the definitive Jobs biography, Becoming Steve Jobs. The fourth estate originated during the French Revolution as a check on the clergy, nobility, and commoners, and evolved in America into a press that sits outside the three branches of government — a concept that only became formalized after the 1920s, largely sparked by Upton Sinclair's exposé of the meatpacking industry. The Alsop brothers — Stewart's grandfather and great-uncle — built their journalistic credibility by taking on Joe McCarthy at the height of his power, which gave them enough reputational armor to withstand the later revelation that they had been informally debriefing the CIA after foreign trips. Russia is neither fully Western nor Eastern — it spans eleven time zones, was shaped by Mongol rule, replaced the Tsar with communism, and at one point sent quiet diplomatic signals about wanting to join NATO, not as a junior member but as a great power on par with the US and China. AI can only build a picture of you from the digital trail you've left behind — and for anyone whose active years predate Gmail, that trail barely exists, making tools like the digital twin app Sentience far less useful for older generations. Journalism and history are fundamentally different disciplines: journalists capture the present moment, while historians piece together the past from whatever fragmentary records survived — a challenge that becomes vivid when trying to reconstruct what Stewart Alsop II actually said in a speech he gave in Soviet-era Moscow. The rationalist movement around figures like Eliezer Yudkowsky, which helped seed effective altruism and shapes thinking at places like Anthropic, may be strong on the technical mechanics of AI but weak on understanding how AI will actually play out in a human world — because humans are not, and have never been, purely rational actors.

  • S1 · E95
    June 25 · 59 min

    Episode #95: Schrodinger's Bubble: Nobody's Keeping Up, And That's Okay

    In this episode of the Stewart Squared podcast, host Stewart Alsop sits down with his father, guest Stewart Alsop II, to tackle corrections from last week's show before diving into the rapid pace of AI development and whether anyone can truly keep up. They explore Brian Chesky's new AI lab venture, Anthropic's controversial Fable release and subsequent restrictions by the US government, and Stewart's increasingly frustrating relationship with what he calls an "abusive superintelligence." The conversation shifts to immersive art experiences as Stewart Alsop II reports from Japan, comparing his visit to TeamLab's digital projection exhibits with his investment in Meow Wolf's physical installations. They discuss the business models behind immersive entertainment, the limits of current AI capabilities (spoiler: AGI definitely isn't here yet), and why FFMPEG might be the unsung hero of modern video software. The episode wraps with reflections on Japan's art island Naoshima and the future of live streaming the podcast. Timestamps 00:00 Welcome and experiment announcement: Stewart introduces a new fact-checking approach for the podcast, explaining how they'll correct previous episodes while maintaining their improvisational conversation style. 05:00 Correcting last week's record: The hosts address three mistakes from the previous episode regarding Brian Chesky staying as Airbnb CEO, Anthropic's revenue numbers, and NVIDIA's history with Apple's Mac computers. 10:00 The impossibility of catching up: Discussion of Stewart II's newsletter concept about falling behind in the AI race, examining Meta and XAI's struggles to compete with leading AI companies despite massive investments. 15:00 Schrodinger's bubble theory: Stewart explores whether we're experiencing a tech bubble, comparing current AI acceleration to past technological shifts and discussing uncertainty around market valuations. 20:00 Abusive superintelligence relationship: Stewart describes his frustrating experience with Anthropic's constant changes, quality degradations, and trust issues while building applications dependent on their AI models. 25:00 Enterprise focus and philosophical concerns: Analysis of Anthropic's shift toward enterprise customers, their cult-like hiring practices, and concerns about effective altruism ideology influencing AI alignment decisions. 30:00 Geographic restrictions and sovereignty: Discussion of Fable's sudden unavailability to non-US citizens, prompting exploration of Chinese AI models as alternatives for maintaining independence. 35:00 Immersive entertainment comparison: Stewart II shares impressions from visiting TeamLab in Tokyo, comparing their digital projection-based experiences with Meow Wolf's physical installations and business models. 40:00 TeamLab versus Meow Wolf analysis: Detailed comparison of how TeamLab uses programmable projections for repeatability while Meow Wolf builds physical environments, discussing advantages and challenges of each approach. 45:00 Business model differences: Exploration of capital costs, repeat visitors, and sustainability challenges between TeamLab's digital flexibility and Meow Wolf's expensive physical build-outs in multiple cities. 50:00 Live streaming ambitions: Stewart reveals plans to livestream future episodes using FFMPEG technology, discussing the technical challenges and open-source philosophy behind modern video streaming infrastructure. 55:00 Japan's art island experience: Stewart II describes visiting Naoshima, an island dedicated entirely to art installations including works by David Hockney and Yayoi Kusama's famous pumpkin sculptures. Key Insights 1. The podcast experimented with a new format of correcting factual errors from previous episodes, including clarifications about Brian Chesky remaining as Airbnb CEO while building a separate AI lab, corrections to Anthropic revenue figures, and historical facts about NVIDIA providing GPUs to Apple products until around 2012-2013. This represents an effort to maintain journalistic accuracy despite the improvised nature of their conversations. 2. A central thesis emerged around the impossibility of catching up in the AI race once a company falls behind. Examples include Meta's struggles despite aggressive researcher hiring and expensive talent acquisition, and XAI renting out unused data center capacity to competitors like Anthropic and Google for billions per quarter, suggesting their product is not achieving comparable usage to competitors despite massive infrastructure investment. 3. The concept of Schrodinger's bubble was introduced to describe the current technological moment, where we exist in an uncertain state between revolutionary transformation and speculative excess. Unlike previous acceleration periods in the 1980s-2000s with personal computers or social media's emergence, this acceleration with AI appears unrelenting, and determining whether we are in a bubble is impossible until the bubble either continues or bursts, creating anxiety and excitement simultaneously. 4. Anthropic faces criticism for degrading service quality and implementing paternalistic guardrails on their Fable model, including downgrading performance in certain domains like biotech and cybersecurity, sometimes without user notification. This approach to AI alignment, rooted in effective altruism philosophy, is viewed as potentially deluded and cult-like, prioritizing enterprise customers over individual users while destroying trust through policies like restricting non-US citizens from accessing certain features. 5. The comparison between immersive entertainment experiences TeamLab in Japan and Meow Wolf reveals fundamentally different business models, with TeamLab using digital projection that can be easily reprogrammed versus Meow Wolf's expensive physical builds. TeamLab likely achieves more repeat business through constantly changing digital experiences, while Meow Wolf struggles with high capital costs and limited reasons for visitors to return, suggesting future convergence between these approaches. 6. Current AI capabilities fall short of artificial general intelligence, as demonstrated by persistent failures to solve complex technical problems like real-time video lip syncing despite access to advanced models like Anthropic's Fable. While AI excels at deterministic software tasks with automated tests, it cannot handle subjective domains requiring taste like video production or immersive experiences, revealing fundamental limitations in current large language models. 7. Open source technology like FFMPEG demonstrates how fundamental video and audio processing capabilities remain available to everyone on a level playing field, with major platforms like YouTube, Netflix, and Rumble all using the same underlying tools. This represents a successful counter-model to proprietary complexity from the 1990s, suggesting opportunities for new competitors to build sophisticated streaming and video capabilities without requiring the resources of established tech giants.

  • S1 · E94
    June 18 · 51 min

    Episode #94: ARM Wrestling: NVIDIA's Quiet Coup Against Intel

    In this episode of the Stewart Squared podcast, host Stewart Alsop speaks with his father Stewart Alsop II, who joins from Tokyo while Stewart broadcasts from Buenos Aires at 5 AM his time. The conversation covers NVIDIA's new Spark chip announcement and its partnership with Microsoft to bring ARM-based processors to Windows PCs, finally allowing Windows to compete with Apple's performance gains from five years ago when they switched to their own ARM-based M-series and A-series chips. They discuss the competitive dynamics between chip manufacturers, the token apocalypse affecting AI coding assistants like Claude and Codex, and how companies like Anthropic are struggling with inference costs while renting data center capacity from SpaceX's underutilized X AI facilities. The discussion also touches on the rise of small models for on-device AI, the dominance of Chinese models in developing markets, SoftBank's ownership of ARM and history of big bets, and how attention and access to insider deals have shaped the AI investment landscape. For more context on Microsoft's strategy, Stewart Alsop II references a Ben Thompson Stratechery interview with Microsoft CEO Satya Nadella that helped clarify how Windows now runs on ARM architecture. Timestamps 00:00 Stewart Alsop welcomes listeners, explains recording at 5 AM his time, 5 PM in Tokyo Japan, discusses NVIDIA's new announcement about processors and chips for Windows computers 05:00 Discussion of ARM architecture versus Intel chips, Apple's competitive advantage using ARM-based M-series processors, how Windows has fallen behind Macintosh in performance capabilities 10:00 NVIDIA positioning new chip as AI-focused but actually ARM-based architecture, Microsoft modifying Windows to run on ARM, multiple manufacturers producing laptops with NVIDIA chips instead of Intel 15:00 Deep dive into ARM licensing model, SoftBank ownership of ARM, how NVIDIA's CPU competes with Intel while Microsoft adapts Windows for ARM architecture 20:00 Intel's competitive position, Microsoft's alliance with NVIDIA, discussion of GPU versus CPU functions, how graphics processing naturally supports training large language models 25:00 Token apocalypse experience with Claude and Codex, rate limiting issues, moving between coding assistants, quality regressions and improvements in different AI coding tools 30:00 Anthropic efficiency improvements with Opus 4.8, competitive dynamics between Claude Code and Codex, strategy of using multiple subscriptions to avoid rate limiting 35:00 Chinese models as workhorses for global users who cannot afford expensive subscriptions, frontier models limited to Google Anthropic and OpenAI, affordability challenges internationally 40:00 Small models running on devices versus cloud-based large models, Apple's WWDC expectations for integrating models on iPhone, personal computing productivity shifts 45:00 SoftBank history with Masayoshi Son making big bets, ARM acquisition rationale, attention-based access to insider deals, comparison to celebrity entrepreneurs gaining investment access 50:00 Historical perspective on insider access to deals and IPOs, closing remarks about continuing conversation from Japan Key Insights 1. Microsoft and NVIDIA announced a new ARM-based processor called Spark that will run Windows, marking a significant shift in the PC market. This represents Microsoft finally moving away from its dependence on Intel chips, similar to what Apple did five years ago when it introduced its M-series chips for Macintosh computers and A-series for iPhones. The development is positioned as an AI chip for marketing purposes, but the real significance lies in the ARM architecture, which NVIDIA has licensed. This alliance between Microsoft and NVIDIA directly challenges Intel's dominance in the PC processor market and could make Windows machines more competitive with Apple's Macintosh in terms of performance and efficiency. 2. The competitive landscape in AI coding assistants has dramatically shifted, with Anthropic's Claude Code releasing version 4.8 that significantly improved code quality and token efficiency. After experiencing severe rate limiting issues in May due to inference capacity constraints, Anthropic made their coding model much more efficient at the token level, allowing users to accomplish more within existing subscription tiers. Meanwhile, OpenAI responded with Codex to compete with Claude Code's success from last December. This competition has created a situation where programmers are now splitting subscriptions between multiple services, paying for both Codex and Claude Code while using Chinese open-source models as fallback options when they hit rate limits. 3. The token apocalypse revealed fundamental business challenges for AI companies as they struggle to balance inference capacity with growing demand. Anthropic had to make difficult decisions to prioritize enterprise customers over individual users, causing noticeable degradation in their chatbot product quality. The company was spending enormous amounts of inference capacity on making conversations feel natural and philosophically relevant, which proved financially unsustainable. Companies like Uber reportedly burned through their entire token budgets in just three months, highlighting how the rush to maximize token usage became a poor metric for actual productivity, falling victim to Goodhart's Law where a measure that becomes a target ceases to be a good measure. 4. The revenue growth projections for Anthropic demonstrate the explosive commercial potential of large language models. The company expected to end 2025 with 9 billion dollars in revenue, but by the second quarter had revised expectations to 50 billion dollars. This astonishing growth comes from companies paying substantial enterprise budgets for AI services. Meanwhile, SpaceX's X AI data center, built rapidly but underutilized due to poor adoption, has been rented out to both Anthropic and Google for approximately 2 billion dollars per month collectively, showing how infrastructure built for one purpose can be repurposed when the original business model fails to generate sufficient demand. 5. SoftBank's strategic bet on ARM five years ago positioned the company at the center of the current processor revolution. Founded by Masayoshi Son, SoftBank has a history of making large, bold investments over four decades, including early deals with Microsoft for software distribution in Japan. The company took ARM private and then public again, with SoftBank retaining majority ownership. This investment proved prescient as ARM's licensing model became increasingly valuable, especially as Apple, NVIDIA, and others adopted ARM architecture for their processors, making it the de facto standard for CPU design across multiple device categories from smartphones to personal computers. 6. The future of AI appears to be splitting between small models running on devices and large frontier models in the cloud. Apple is expected to announce at WWDC its integration of Google models on the iPhone, utilizing small models that can run locally on the device for personal productivity tasks like calendar and email management, while connecting to cloud-based large language models for more complex operations like programming. This hybrid approach addresses both privacy concerns and cost efficiency, as running everything through cloud-based large language models proves financially unsustainable for everyday personal computing tasks. The industry consensus currently recognizes only three companies as leaders in frontier models: Anthropic, OpenAI, and Google. 7. Chinese AI models are emerging as the workhorses for global markets due to a...

  • S1 · E93
    June 11 · 53 min

    Episode #93: Too Big to Question: SpaceX, Wall Street, and the End of Accountability

    In this episode of the Stewart Squared podcast, host Stewart Alsop and guest Stewart Alsop II tackle the explosive SpaceX IPO, conflict of interest in politics and finance, and whether we're heading toward economic collapse or the singularity. The conversation kicks off with them acknowledging they had to restart recording after getting into a heated argument about whether Trump's stock trading and Nancy Pelosi's husband's trades fall into the same category of insider dealing—though neither technically qualifies as illegal insider trading. From there, they dig into the mechanics of the SpaceX IPO, questioning how Elon Musk convinced major banks like Goldman Sachs and Morgan Stanley to support a staggering $1.75 trillion valuation despite the company reporting nearly $5 billion in losses against $18.7 billion in revenue. Stewart II, who actually read the 300-page S-1 prospectus (unlike most people), explains how this IPO could fail and compares it to the infamous WeWork collapse. They explore the manual process still involved in IPOs, the role of stock exchanges from the Dow to NASDAQ to the new Texas Stock Exchange, and how Trump has concentrated executive power in ways that echo—and pervert—Teddy Roosevelt's use of executive orders. The discussion touches on reserve currencies, Argentina's economic history, cryptocurrency's death as a decentralized ideal, and whether the singularity is real or just conspiracy fantasy embraced by wealthy tech elites. Timestamps 00:00 Stewart Squared podcast begins with revealing an argument about Trump and Nancy Pelosi both doing insider trading though it's not technically illegal insider trading 05:00 Discussion shifts to insider trading history from the Great Depression era and how current rules no longer work effectively with both politicians stretching ethical boundaries thin 10:00 SpaceX IPO prospectus analysis begins with focus on Elon Musk's control and conflicts of interest as banks go along with questionable trillion dollar valuation for massive fees 15:00 Investment banking history explored from boutique banks in seventies taking startups public to Internet bubble abuses and evolution through social media crypto and AI eras 20:00 Stock exchanges worldwide discussed including NASDAQ origins in 1971, New York Stock Exchange history, and newer Texas stock exchange where Elon sells shares with fewer reporting rules 25:00 Chevron principle explanation showing how Trump gathered executive power while claiming to fight deep state creating ironic situation of doing more executive overreach not less 30:00 US dollar reserve currency status threatened by massive national debt and interest payments now consuming thirty percent of federal budget with neither party willing to balance accounts 35:00 IPO mechanics and pricing discussed with SpaceX seeking up to two trillion valuation though market expects between one trillion and 1.6 trillion based on Polymarket betting 40:00 Risk factors in SpaceX prospectus examined including losses of 4.9 billion against 18.7 billion revenue creating outrageous 300x price to sales ratio with Elon controlling 85 percent voting 45:00 Argentina economic crisis comparison drawn from 1960s through 2001 Corralito when peso devalued from one-to-one with dollar to one-to-four overnight destroying savings 50:00 Singularity discussion concludes episode calling it conspiracy fantasy while drawing parallels between Theodore Roosevelt's executive orders for public good versus Trump's for personal profit Key Insights 1. The discussion reveals a fundamental transformation in how stock markets and Initial Public Offerings function compared to historical norms. The SpaceX IPO represents an extreme example of this shift, with Elon Musk essentially controlling the entire process including valuation, pricing, and disclosure while investment banks like Goldman Sachs, Morgan Stanley, and JPMorgan simply comply because of the massive fees involved. The IPO aims to raise seventy-five billion dollars at a valuation approaching one point eight trillion dollars, despite the company reporting losses of four point nine billion dollars against eighteen point seven billion in revenue, creating a price-to-sales ratio around three hundred times, which defies traditional financial metrics that would normally support such a valuation. 2. The conversation illuminates how conflicts of interest have become normalized at the highest levels of American finance and government. Trump is described as one of the most active stock market investors while serving as president, with correlations noted between his trades and policy announcements, yet this occurs in an environment where regulatory mechanisms no longer effectively constrain such behavior. The traditional checks and balances that prevented insider trading and conflicts of interest have been stretched so thin that nobody can agree on what constitutes inappropriate behavior anymore, creating a system where all rules have become negotiable for those with sufficient power and influence. 3. The decline of traditional IPO processes reflects broader systemic changes in American capitalism. In the nineteen seventies and eighties, boutique investment banks would take startup companies public when they had thirty to fifty million in revenue at reasonable valuations, providing opportunities for companies to access public markets relatively quickly. That system was abused during the Internet bubble of the nineties, leading to companies going public and then declaring bankruptcy within months. Since then, the market has experienced successive bubbles in social media, crypto, and AI, with each cycle becoming progressively more detached from fundamental business metrics and increasingly difficult to distinguish sustainable businesses from speculative ventures. 4. The role of stock exchanges has evolved significantly, with the NASDAQ emerging in 1971 specifically to serve technology companies while the New York Stock Exchange dates back to the 1890s. The conversation reveals that SpaceX is being included in the Dow Jones index immediately upon going public, rather than waiting the typical six months, and that Musk is also selling shares on the newly created Texas Stock Exchange where regulations are less stringent. This fragmentation of markets and willingness to bend traditional rules for high-profile offerings demonstrates how institutional guardrails have weakened, with exchanges competing for prestigious listings by offering more favorable terms rather than maintaining consistent standards. 5. The discussion of reserve currency status reveals existential risks facing the American economy. The United States has maintained the dollar as the global reserve currency, which allows the country to borrow its way out of trouble because all other currencies are indexed to it. However, this system is being abused through massive national debt where interest payments now consume roughly thirty percent of the federal budget. Neither Republicans nor Democrats are willing to bring operating accounts back into balance, and there are now situations where countries trade currencies without reference to the dollar. If the United States loses reserve currency status, the country would face an Argentina-like scenario of economic collapse. 6. The comparison between current conditions and historical economic crashes provides important context for understanding present risks. The speakers identify that the 2008 crash was triggered by real estate, the 2001 crash by the Internet bubble, and 2020 by the pandemic, but the trigger for the next crash cannot be predicted in advance. What makes the current situation particularly concerning is that multiple sectors appear overvalued simultaneously, with unsustainable practices across technology, finance, and government spending. The feeling expresse...

  • S1 · E92
    June 4 · 24 min

    Episode #92: The $1.75 Trillion Bet: What WeWork Taught Us About the SpaceX IPO

    On this episode of the Stewart Squared podcast, host Stewart Alsop speaks with his father Stewart Alsop II about the SpaceX IPO and whether such a massive public offering could actually fail. Stewart Alsop II published his analysis just fifteen minutes before recording at sallsop.substack.com, questioning the logic behind the $1.75 trillion valuation and $75 billion raise, especially given that the company loses nearly $5 billion annually. The conversation ranges from the mechanics of IPOs and the SEC approval process to the only recent failed IPO (WeWork), SPACs versus traditional public offerings, the iron triangle of regulators and business interests, and comparisons between political figures' investment track records. Stewart Alsop II draws on his experience living through decades of Bay Area politics and business while analyzing whether institutions will actually buy into what he describes as a bet on Elon Musk rather than traditional fundamentals. Timestamps 00:00 Stewart introduces the episode topic returning to SpaceX IPO discussion and what he learned writing his article about IPO failures 05:00 Discussion of how IPO conspiracy works between SEC regulators bankers and entrepreneurs creating iron triangle relationships that rarely result in failures 10:00 WeWork becomes example of rare IPO failure when institutions refused to buy shares despite SEC approval and banker support 15:00 Examination of Trump's public transparency about money-making versus traditional banana republic secrecy and oligarch networks 20:00 Debate over World Liberty Financial investments and whether Trump family portfolio signals SpaceX IPO success potential 25:00 Heated disagreement about Nancy Pelosi's husband's stock trading and conspiracy theories before ending the episode Key Insights 1. An IPO can fail after the S-1 filing is published, though it has only happened once in recent memory with WeWork. When Adam Neumann pushed bankers to file WeWork's S-1, institutional investors reviewed the disclosed information and refused to buy the stock, preventing the company from going public through the traditional IPO process. This demonstrates that while the SEC, bankers, and company founders may all approve of an offering, the ultimate gatekeepers are the institutional investors who actually purchase the shares. 2. The SpaceX IPO represents an unusual situation where the company seeks a valuation of approximately 1.75 trillion dollars while only raising 75 billion dollars, representing roughly 2% of the company. This creates a challenging situation for potential investors because the upside is limited—for investors to make significant returns, SpaceX would need to become worth more than Apple, Google, or NVIDIA, all of which have twenty-year histories as public companies. This raises serious questions about the rational investment case for institutional buyers. 3. Investment bankers have strong financial incentives to push IPOs through to completion, as they receive approximately 6% of the proceeds. In the SpaceX case, this would amount to 6% of 75 billion dollars. This creates a structural problem in the IPO process where bankers may not adequately filter out questionable offerings, relying instead on the SEC approval process and institutional investor appetite to serve as quality controls. 4. Elon Musk has consolidated multiple companies into SpaceX before proposing to go public, including X AI and the company formerly known as Twitter, in addition to the core SpaceX rocket business and Starlink. While SpaceX itself generates about 4.5 billion in revenue and Starlink generates 11.5 billion in revenue growing at 50% annually, the company is currently losing almost 5 billion dollars per year. This makes the offering a bet on much more than just the space business, and Musk will control 85% of voting shares. 5. SPACs, or Special Purpose Acquisition Companies, represent an alternative path to going public that bypasses the traditional IPO process. These shell companies go public at 10 dollars per share without having an actual operating business, then search for a private company to merge with. WeWork eventually went public through a SPAC after its traditional IPO failed, though the company later went bankrupt. Most SPACs, approximately 95%, trade below their original value, making them generally problematic investment vehicles. 6. Elon Musk has successfully transformed two major industries through Tesla and SpaceX, which distinguishes him from other entrepreneurs like Adam Neumann who had not proven themselves before WeWork. Tesla proved the viability of electric cars, challenged dealer rules to sell directly to customers, built charging networks, and manufactured batteries at unprecedented scale. Similarly, SpaceX developed the reusable Falcon 9 rocket and built Starlink into a business three times larger than the rocket business itself, though Musk nearly went bankrupt three times in the process. 7. The traditional IPO process involves an iron triangle between the SEC regulators, investment bankers, and company founders, with institutional investors serving as the final check on whether an offering succeeds. Approximately 98% of IPO shares are purchased by institutions rather than individual investors. The SEC requires companies to publish an S-1 disclosure document revealing all company details, and if this document passes SEC review, bankers then attempt to sell shares to institutional investors who make the ultimate decision about whether to participate.

  • S1 · E91
    May 28 · 46 min

    Episode #91: The $1.5 Trillion Question: Why SpaceX's IPO Math Doesn't Add Up

    In this episode of the Stewart Squared podcast, host Stewart Alsop and his father Stewart Alsop II dig into the major AI and tech IPOs hitting the market, with SpaceX leading the charge at a controversial $1.5 trillion valuation despite just $20 billion in revenue. They break down how SpaceX's massive S-1 filing (so big it crashed Claude's context window) reveals a company now bundling together Starlink, rocket launches, X (Twitter), and the struggling xAI/Grok business—with key researchers having already jumped ship after getting their SpaceX stock. The conversation covers Anthropic's explosive revenue growth (projecting $10 billion in Q2 alone) and their smart move renting Musk's underutilized data center for $1.25 billion, OpenAI's pending IPO, Apple's quiet but strategic AI approach using on-device models and partnering with Gemini instead of OpenAI, and why the institutional investors might balk at SpaceX's aggressive pricing when the IPO drops on June 12th. Stewart II shares his contrarian take: he'd never touch SpaceX stock at this valuation but is seriously considering Anthropic, while explaining the arcane details of revenue recognition, vesting schedules, and why Elon Musk's singular track record lets him operate by different rules than any other CEO. Timestamps 00:00 Welcome and SpaceX IPO discussion begins, exploring the $20 billion valuation and mathematical implications of the massive offering 05:00 Anthropic renting Musk's data center for over a billion monthly while Grok struggles, researchers leaving xAI after receiving SpaceX stock 10:00 Institutional investors may decline SpaceX shares at ridiculous valuation compared to Apple's $400 billion revenue and NVIDIA's profitability 15:00 Apple's on-device AI strategy with small models and Gemini integration while Musk fails in foundation models 20:00 Revenue recognition differences between companies, Anthropic projecting $10 billion quarterly revenue with conservative accounting practices 25:00 SpaceX revenue breakdown showing Starlink at $11 billion dominating over rocket business, Twitter and xAI tucked into valuation 30:00 Comparing SpaceX's $20 billion revenue to Apple's $400 billion while discussing material disclosure requirements in IPO filings 35:00 Musk's singular achievement changing space and car industries, earning unprecedented valuation despite rational market concerns 40:00 Argentine politics and Milei's challenges, parallels to Trump's midterm influence and Peter Thiel's strategic positioning 45:00 Final thoughts on IPO opportunities, avoiding SpaceX at current valuation while considering Anthropic's rapid growth potential Key Insights 1. SpaceX is going public at a 1.5 trillion dollar valuation while generating only 20 billion in revenue, creating significant concerns about whether the IPO will succeed. The company is attempting an unusually fast timeline from S-1 filing on May 20th to going public on June 12th, bypassing the typical two month roadshow process. There is a real possibility the offering could fail because institutional investors who must buy 70% of the shares may decline at this valuation, seeing no path for the stock to appreciate further. 2. The valuation appears disconnected from fundamentals when compared to companies like Apple with 400 billion in revenue worth 4 trillion or NVIDIA with 85 billion in revenue worth 5 trillion. SpaceX would need to grow revenue from 20 billion to potentially 100 billion and achieve profitability to justify even being in the multi-trillion dollar valuation range. The aggressive pricing likely comes from Musk himself rather than the investment banks, as he controls the process with his ownership structure. 3. Anthropic is experiencing explosive revenue growth, jumping from 4 billion in one quarter to a projected 10 billion in the second quarter, putting them on track for 40 to 50 billion in annualized revenue. Most remarkably, they claim they will be profitable in the second quarter, which would be unprecedented for a foundation model company. Their strategic deal to rent Musk's underutilized data center for 1.25 billion monthly solved their infrastructure problems while giving Musk revenue to cover his failed Grok investment. 4. Elon Musk consolidated multiple companies including Twitter, xAI, SpaceX and Starlink into one entity still called SpaceX, creating a complex conglomerate that will be difficult for investors to evaluate. The xAI portion has essentially failed as a foundation model competitor, with dozens of researchers leaving after the merger gave them valuable SpaceX stock as an exit. Twitter contributes roughly 2 billion in revenue, the rocket business does 4 billion, but Starlink is the real driver at 11 billion and growing rapidly. 5. Apple has been quietly working on small on-device AI models embedded in their operating systems rather than pursuing foundation models, and they will likely deliver on their 2024 promises using Google's Gemini instead of OpenAI. This strategic approach of focusing on practical on-device capabilities while partnering for cloud capabilities may prove more successful than trying to build their own foundation model. The company avoided the mistake of announcing capabilities before they were ready, then pragmatically adjusted their approach. 6. Once you fall behind in the foundation model race, you cannot catch up, which explains why both Musk with Grok and Zuckerberg with Meta have struggled despite massive investments. The leaders like Anthropic and OpenAI have such strong momentum and embedded positions that competitors cannot overcome the gap. This dynamic is similar to how Palantir embedded itself so deeply in government and commercial customers before LLMs that they remain entrenched despite new AI capabilities. 7. The simultaneous IPOs of SpaceX, OpenAI and Anthropic represent different investment propositions, with SpaceX being personality and potential driven, OpenAI having revenue recognition questions, and Anthropic showing the strongest fundamentals with explosive growth and a path to profitability. All three will have founder-controlled voting structures similar to Meta where Zuckerberg has 60% control, allowing these leaders to pursue long-term visions regardless of public market pressures. The timing is largely coincidental rather than coordinated, driven by each company's specific capital needs and market conditions.

  • S1 · E90
    May 21 · 52 min

    Episode #90: Nobody Knows What Software Is Worth Anymore

    In this episode of the Stewart Squared podcast, host Stewart Alsop II connects from Tangier, Morocco while his son Stewart Alsop III digs deep into the technical challenges of building video conferencing software, specifically tackling the notorious lip sync problem that's consumed his last two months. The conversation moves from mutation testing and DevOps to exploring the future of software consulting, examining why Silicon Valley has long held a visceral distrust of consultants while contractors thrive, and what AI-powered development means for how software gets built and sold in the coming years. Stewart III shares his journey from "vibe coding" to implementing scientific methods in his development process, while his father draws on decades of experience as both a journalist and investor to contextualize the shifting landscape of enterprise software, touching on everything from the rise of SaaS to why companies like Riverside raised $80 million while Stewart III builds competing technology solo in his head. Timestamps 00:00 Welcome from Morocco, Stewart Senior joins from Tangier with Middle Eastern backdrop, Stewart Junior deep in AI development learning mutation testing, integration tests, unit tests, red to green testing 05:00 Discussion of vibe coding evolution to scientific method coding, working on lip sync white whale problem for two months, building pipeline from recording to post-production using FFMPEG diagnostics 10:00 Explanation of how recording works with separate audio and video streams, discovery that browser clocks using tiny crystals don't keep accurate time, learning about MediaRecorder API versus WebCodecs advantages 15:00 Debate about competing with Riverside's 80 million dollar funding, discussion of building specialized software versus SaaS products, exploring turnkey podcasting solutions and business models 20:00 Deep dive into consultancy business model, Stewart Senior's visceral hatred of consultants, discussion of business school graduates becoming consultants or bankers, Microsoft's deliberately small consulting practice 25:00 Exploration of conflict of interest in journalism and investing, disclosure requirements, comparison to New York Times OpenAI lawsuit, discussion of father's unpaid consulting role in DC power centers 30:00 History of consultancies like Arthur Andersen and PricewaterhouseCoopers, role in mergers and acquisitions, example of David Ellison buying Paramount and pursuing Warner Brothers Discovery 35:00 Difference between contractors and consultants, discussion of outsourcing to India, Cloud Factory in Nepal, Ronald Coase economics, Infosys as first big software engineering consultancy 40:00 Stewart Junior's ability to understand code concepts without reading code, using scientific method and chaos monkey development, Netflix streaming techniques, debugging through sufficient motivation 45:00 Sales challenges and negotiation skills in family, working with mentor Zavant on sales frameworks, generosity versus transactional relationships, Turkish bazaar negotiation culture comparison 50:00 Discussion of value creation and belief in sellability, the 80/20 rule of product completion, Adam Neumann and Travis Kalanick examples, Elon Musk as builder not salesman creating entire systems Key Insights 1. The challenge of solving technical problems reveals the importance of understanding methodologies over mastering code itself. Stewart Alsop III spent two months wrestling with a lip sync problem in his video recording system, learning about mutation testing, integration tests, and DevOps along the way. The key insight is that he does not need to read or write code directly anymore. Instead, he needs only a conceptual understanding of frameworks like the scientific method or chaos engineering to direct AI systems to solve complex technical problems. This represents a fundamental shift where domain knowledge and problem articulation matter more than programming expertise. 2. Modern video conferencing systems create synchronization challenges because different computers use tiny crystals to keep time, but these crystals do not maintain perfect accuracy, especially when network conditions fluctuate. The problem is not simply about recording separate audio and video streams and reassembling them. Instead, systems create containers with audio and video together while also recording separate audio tracks, and all these different clocks drift apart from each other. This explains why lip sync issues plague even well funded platforms like Riverside, and why solving this problem requires sophisticated diagnostic systems and conversion pipelines using tools like FFMPEG. 3. The evolution of software business models reflects changing technological constraints and market conditions. In the 1990s and early 2000s, software was sold as one time purchases, often on physical media like cartridges or floppy disks. The shift to Software as a Service in the 2010s happened because it was considered better for customers who did not have to pay large upfront fees and because cloud infrastructure made it feasible. Now, with AI enabling individuals to build complex software themselves, we may be entering another transition period where the SaaS model itself becomes obsolete, though what will replace it remains unclear. 4. Programming represents the first domain where artificial general intelligence has effectively arrived because programming consists entirely of text. Unlike domains involving physical manipulation or subjective judgment, code can be completely represented in language, and decades of open source code provide massive training datasets. This explains why tools like Claude have become so powerful so quickly in programming contexts, and why Anthropic claims that most of its models are now generated by AI systems themselves. The recursive nature of AI writing code to improve AI represents a fundamental breakthrough that does not yet exist in other domains. 5. Consultancies emerged to solve problems that companies could not efficiently solve themselves, but their value proposition is eroding. Large consulting firms like the Big Seven accounting firms grew powerful by integrating complex enterprise software and managing mergers and acquisitions. However, as software becomes easier to build and modify through AI, and as the difficulty of integration decreases, the justification for expensive consultancies diminishes. The antipathy toward consultants in Silicon Valley stems from a belief that they represent companies paying others to think for them rather than developing internal capabilities, and this critique becomes more valid as technical barriers fall. 6. The distinction between contractors and consultants matters for understanding business models and value creation. Contractors are individuals or small teams hired for specific projects who sell their labor directly. Consultancies are businesses built around winning large contracts and then deploying teams to execute them, often with substantial markup. The emergence of platforms like Upwork and the phenomenon of outsourcing to places like India, Nepal, and Kenya created hybrid models where individual profiles often mask small consultant operations. Understanding these distinctions helps clarify what kind of business model makes sense for someone developing new technical capabilities. 7. Believing in the value of what you create is a prerequisite for being able to sell it, and products must be truly finished before they have sellable value. The last twenty percent of any project, whether writing, programming, or product development, represents the hardest work because it involves transforming something functional into something polished and complete. Until the lip sync problem is definitively solved, the video recording system remains a prototype rather ...

  • S1 · E89
    May 14 · 1 hr 7 min

    Episode #89: Vibe Engineer Meets Venture Capitalist: A Father-Son Dispute About the Future

    In this episode of Stewart Squared, host Stewart Alsop sits down with his father, Stewart Alsop II, for a wide-ranging conversation that moves from the technical to the historical to the financial. The two kick things off with Stewart's self-proclaimed evolution from "vibe coder" to "vibe engineer," as he tackles the tricky challenge of audio and visual sync in his own custom podcast recording software, positioning it as a direct competitor to platforms like Riverside.fm and Squadcast. From there, they get into a business breakdown of OpenAI and Anthropic, debating whether Claude's recent stumbles are a blip or a sign of deeper trouble, and what an IPO would actually mean for both companies as they look to compete with the big players. The conversation winds through a rich history of personal computing — from Mosaic and Netscape to PageMaker and the LaserWriter, desktop publishing, the browser wars, and how Windows 95 and the early internet reshaped everything — before landing on the turbulent state of the airline industry, the fallout from the Strait of Hormuz blockade, and what the collapse of Spirit Airlines says about fragile business models. Timestamps 00:00 - Stewart introduces vibe engineering, tackling audio-visual sync problems while others debate AI coding tools. 05:00 - Deterministic vs probabilistic software discussed, with Stewart building real engineering skills through coding challenges. 10:00 - Browser history explored, from Mosaic origins at University of Illinois to Netscape's proprietary commercialization. 15:00 - Adobe Flash wars with Steve Jobs examined, leading into desktop publishing revolution with PageMaker and LaserWriter. 20:00 - PostScript origins at Xerox PARC discussed, Adobe founders transforming page composition from compositors to editors. 25:00 - Kinkos, Windows vaporware, and personal computing evolution from 1985 through Windows 95 emergence. 30:00 - Information Superhighway era examined, Netscape on Windows 95 driving personal computer mainstream adoption. 35:00 - Claude versus Codex battle analyzed, Anthropic's trust erosion among engineers and Silicon Valley insider bubble. 40:00 - OpenAI versus Anthropic growth metrics compared, IPO strategies and public market ambitions dissected. 45:00 - Stock fundamentals explained through Tesla versus traditional automakers, quarterly earnings disclosure requirements. 50:00 - Airline complexity breakdown, Spirit Airlines collapse tied to jet fuel hedging failures post-Iran blockade. 55:00 - New capitalism emerging through AI, IPO mechanics enabling OpenAI and Anthropic to compete with tech giants. 01:00:00 - Meta, Apple, Microsoft AI strategies compared, Chinese model competition driving Anthropic's existential decisions. 01:05:00 - Surveillance states, sovereign nations, and India versus small countries as future nonaligned powers debated. Key Insights 1. There is a meaningful distinction emerging between types of AI-assisted builders. Actual engineers use AI tools to boost productivity while still understanding code. Vibe coders use prompt engineering to build things without formal training. And then there are people who have no interest in building software at all because they simply do not need to. 2. Deterministic software is fundamentally different from probabilistic AI outputs. While the current hype around AI agents and markdown-based workflows is real, the underlying products are often insecure and unreliable. Building deterministic software first and layering in AI agents later is a more stable and trustworthy approach. 3. Desktop publishing in the mid-1980s was a landmark moment in personal computing. The combination of the Apple Macintosh, PageMaker, and the LaserWriter printer transferred control of page composition from professional compositors to individual editors and writers, democratizing the ability to produce print materials. 4. The browser wars of the 1990s, particularly Netscape running on Windows 95, marked the moment when the personal computer became meaningful to ordinary people. Before that, roughly a decade passed where developers and companies were still figuring out how operating systems, platforms, and application development were supposed to work together. 5. The MediaRecorder API is a significant but underappreciated limitation in modern browser development. Because Safari does not support it in the same standardized way as Chrome and Chromium-based browsers, many podcast and recording platforms are effectively locked to Chrome, creating an opening for alternative technical approaches. 6. Going public through an IPO gives companies like OpenAI and Anthropic access to capital at a scale that private fundraising cannot easily match. It also imposes mandatory quarterly financial disclosures, which means the public will finally be able to see actual revenue, spending, and growth figures rather than relying on perception and valuation claims. 7. Airlines represent one of the most operationally complex businesses in existence, involving gate leases, dynamic ticket pricing, fuel costs, crew logistics, and massive debt structures. The sudden spike in jet fuel prices following the US blockade of the Strait of Hormuz exposed airlines that had not hedged their fuel costs, contributing directly to Spirit Airlines going out of business.

Showing 1–20 of 27 episodes