Anthropic S-1 Shows $67B Run Rate, Founders Keep 50.1% Vote
TBPN
Anthropic achieved operating profitability with only $482 million in first-half capital expenditures by renting rather than owning its compute infrastructure, while isolating 50.1% voting control in a zero-equity golden share held by its seven founders.
Anthropic filed its S-1 prospectus revealing a $67.2 billion annualized revenue run rate, positive operating margins (+19%), and $131.7 billion in cash and short-term investments after burning under $10 billion lifetime. (0:00)
Anthropic's long-term cloud and hardware commitments reach $637 billion across Google Cloud, AWS, Microsoft Azure, Broadcom, and SpaceX AI, reflecting a pure compute-rental operational strategy. (4:08)
A bespoke corporate governance structure grants Anthropic's seven co-founders 50.1% voting control via a single Class F 'golden share' held in Founder LLC, decoupling absolute board control from economic equity ownership. (13:38)
Security researchers uncovered tracking mechanisms inside Samsung Smart TV software, including a region-specific convolutional neural network monitoring viewing habits and references to facial recognition code. (30:00)
Detailed notes
## Background
Published on YouTube on October 26, 2026, by the daily tech talk show TBPN (hosted by John Coogan and Jordi Hays), this 'Diet TBPN' episode delivers a 30-minute recap of major tech headlines. According to the page description, the episode focuses on Anthropic's historic S-1 IPO filing alongside growing spyware allegations surrounding Samsung smart TVs and Instagram's expansion into long-form video.
## TL;DR
Hosts John Coogan and Jordi Hays dissect Anthropic's landmark S-1 IPO prospectus, examining its $67.2 billion revenue run rate, $637 billion in multi-year compute commitments, and a unique Class F 'golden share' that secures 50.1% voting control for its seven founders. The hosts also touch upon model distillation challenges, seasonal developer usage dips, and a recent white-hat AI agent breach of Anthropic's internal communications. Turning to consumer technology, they criticize pervasive living-room tracking and neural network bloatware uncovered in Samsung smart TVs before evaluating Instagram's push into long-form horizontal video to rival YouTube.
## Atmosphere & style
The show carries an energetic, tech-native conversational rhythm, blending irreverent banter—such as tying SEC filings to a holiday countdown—with sharp financial analysis. Coogan and Hays bounce observations off each other with relaxed chemistry, transitioning effortlessly from parsing multi-trillion-dollar cap tables and venture windfalls to voicing collective consumer frustration over sluggish, ad-laden smart televisions. It feels like an insider-informed, fast-moving morning briefing that treats monumental Silicon Valley developments with equal parts technical curiosity and dry comedic wit.
## Bulletin
- [00:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=0s) Anthropic filed its S-1 prospectus revealing a $67.2 billion annualized revenue run rate, positive operating margins (+19%), and $131.7 billion in cash and short-term investments after burning under $10 billion lifetime.
- [04:08](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=248s) Anthropic's long-term cloud and hardware commitments reach $637 billion across Google Cloud, AWS, Microsoft Azure, Broadcom, and SpaceX AI, reflecting a pure compute-rental operational strategy.
- [13:38](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=818s) A bespoke corporate governance structure grants Anthropic's seven co-founders 50.1% voting control via a single Class F 'golden share' held in Founder LLC, decoupling absolute board control from economic equity ownership.
- [30:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1800s) Security researchers uncovered tracking mechanisms inside Samsung Smart TV software, including a region-specific convolutional neural network monitoring viewing habits and references to facial recognition code.
## Key takeaways
- [00:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=0s) Anthropic's S-1 filing reports $67.2 billion in annualized revenue, up 45% quarter-over-quarter, with $131.7 billion in liquid reserves on hand.
- [04:08](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=248s) Unit economics improved dramatically as compute costs dropped from 72% to 48% of revenue, flipping operating margins from -13% to +19%.
- [04:08](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=248s) Rather than building proprietary data centers, Anthropic relies on a multi-year compute-rental backlog totaling roughly $637 billion across major hyperscalers.
- [04:08](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=248s) Customer expansion metrics show a 900% net revenue retention rate among enterprises spending over $100,000 annually, with half of the Fortune 500 currently on board.
- [10:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=600s) Terms of service violations, including bot-driven model distillation farms and cybercrime, account for an estimated 5% to 7% of Anthropic's top-line ARR.
- [10:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=600s) Developer API consumption exhibits distinct seasonal volatility, typically dipping by 20% during the final two weeks of December.
- [13:38](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=818s) Anthropic created a single Class F 'golden share' held by Founder LLC that wields 50.1% majority voting power over director elections and core stockholder decisions.
- [13:38](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=818s) The Founder LLC vehicle allows the seven co-founders to vote as a unified monolithic bloc, preventing external observers from seeing internal dissenting votes.
- [16:30](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=990s) Because the Class F share carries no economic rights, the co-founders can theoretically liquidate equity without surrendering corporate control.
- [20:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1200s) The company's planned public listing targets a valuation between $1.5 trillion and $2 trillion under the ticker symbol ANT.
- [20:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1200s) Early backer Spark Capital committed a total of $777 million across Series C through H, positioning its stake to be worth approximately $18 billion at a $2 trillion valuation.
- [25:12](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1512s) A white-hat security team breached Anthropic's internal Slack systems in under 12 hours using AI agents, highlighting national infrastructure vulnerabilities.
- [25:12](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1512s) Debate over a theoretical $20 billion federal bug bounty weighs the benefit of securing critical infrastructure against the chaos of incentivizing untrained actors.
- [28:57](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1737s) Smart TV manufacturers face consumer backlash for bundling intrusive advertising and tracking networks into display panels instead of offering high-end dumb displays.
- [30:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1800s) Analysis of Samsung Smart TV software revealed local neural networks tracking specific app choices, ad skips, and living room facial recognition cues.
- [31:15](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1875s) Instagram is testing horizontal long-form 'Features' and series hubs to capture smart TV streaming and reduce context-stripping outrage from viral short clips.
The show opens with the hosts declaring Anthropic’s IPO filing the single biggest story of the day—second only to their ongoing holiday tracker. John Coogan and Jordi Hays joke about their running bit of connecting arbitrary tech trivia to the Christmas countdown: with exactly 59 days until Christmas, they playfully wonder if Anthropic is somehow 59 years old, recalling an earlier gag where they noted Peter Thiel happened to be 59 (his birthday having passed two weeks prior) and Jeff Bezos was 62 on day 62. Hays jokes that dropping the S-1 filing exactly 59 days out must contain a subliminal message, urging listeners to hit `Command+F` or turn directly to page 59 of the filing. Looking at page 59, the boilerplate text dryly warns: “We may not have the ability to raise the funds necessary to repay the principal and accrued interest of our outstanding...” showing it is just standard SEC disclosure risk language.
Turning to the actual figures in the massive S-1 prospectus, the hosts unpack headline numbers that frame the global race for AI capital. While *The Wall Street Journal* had speculated about an eventual $100 billion raise, *Bloomberg* framed the filing as Anthropic launching an IPO in a global race for AI cash. Yet, as the hosts observe, Anthropic is already sitting on staggering sums of money:
- **Annualized revenue:** $67.2 billion, surging 45% quarter-over-quarter despite operating at massive enterprise scale.
- **Cash and equivalents:** $131.7 billion on hand, bolstered by bringing in $110.8 billion during the first half alone.
- **Capital efficiency:** Early market commentary hailed Anthropic as potentially the most capital-efficient hyper-growth tech firm in history.
The hosts highlight how the market narrative shifted from the earlier leak, which had fixated on an apparent $80 billion accounting charge. Rather than taking twelve hours for venture capitalists to draft reaction memos, modern analysts deployed Claude and ChatGPT to parse the filing instantly. Jordi Hays remarks that the sentiment represents a definitive market beat, noting: “It feels like this beat expectations. Totally. It didn’t like surprise, oh, they’re actually doing 100 billion in revenue or something. But I mean, really, really solid people haven’t found anything that’s like, oh, I didn’t understand that the business was actually shaped a very different way or something like this.” Coogan jokingly quotes Ayal Harry’s quip that “you know it’s a decent S-1 when Ed Zitron hasn’t tweeted anything,” imagining tech critics fruitlessly searching `Command+F` for words like "bad," "sad," or "problem." With the NASDAQ gaining 1%, the hosts see the broader AI trade turning back on after a period of heavy selling across higher-beta assets like the neo-clouds.
---
[04:08](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=248s) Capex, Hyperscaler Commitments, and Expanding Margins
Delving into the operational mechanics behind the top-line numbers, the hosts point out a striking paradox in Anthropic’s balance sheet: despite holding over $130 billion in cash and raising over $110 billion in H1, Anthropic spent just $482 million on capital expenditures (CapEx) in the first half of the year. Instead of building and owning data centers or custom supercomputing clusters, Anthropic has pursued a pure compute-rental strategy. While an aggregate cloud commitment backlog of roughly $637 billion raised initial eyebrows, Coogan and Hays contextualize the total as a manageable multi-year pipeline spread across major hyperscalers, representing roughly 10x current run-rate revenue stretched across several years through 2033 or 2036.
```mermaid
flowchart TD
accTitle: Anthropic Long-Term Compute Commitments
accDescr: Multi-year compute and cloud spending commitments across hyperscalers and hardware providers.
subgraph Commitments["Multi-Year Compute Backlog ($637B Total)"]
GC["Google Cloud: $114.5B (thru July 2033)"]
AWS["Amazon AWS: $110B (thru 2036)"]
AZ["Microsoft Azure: $45B (thru 2033)"]
SPX["SpaceX AI: up to $120B max"]
BCM["Broadcom: $161B"]
OTH["Lambda / Others: $42B - $45B"]
end
```
The specific breakdown of Anthropic’s long-term infrastructure commitments details:
- **Google Cloud:** $114.5 billion minimum commitment running through July 2033.
- **Amazon Web Services (AWS):** $110.0 billion commitment extending through 2036.
- **Microsoft Azure:** $45.0 billion minimum spend running through 2033.
- **SpaceX AI:** Up to a maximum of $120.0 billion.
- **Broadcom:** $161.0 billion in compute infrastructure agreements.
- **Lambda:** $42.0 billion (alongside other agreements around $45 billion).
The hosts underscore that unit economics and gross margins saw substantial improvements over the period. Compute costs as a share of revenue—effectively the company's cost of goods sold (COGS)—fell precipitously from 72% down to 48%, showing rapid efficiency gains quarter-over-quarter as infrastructure usage stabilized. Operating margins flipped into positive territory, climbing from -13% to +19%.
Customer expansion metrics showed equal velocity. Anthropic reported more than 1,000 customers spending over $1 million annually and over 10,000 enterprise customers spending more than $100,000, achieving a net revenue retention (NRR) rate of 900% on those tiers. The hosts explain that accounts which integrated Claude APIs as minor pilots in 2024–2025 transitioned into the "token-maxing era," resulting in roughly a 10x spend expansion as AI usage moved from lone engineering early adopters to company-wide deployments across operations and marketing.
The disclosure that exactly 50% of the Fortune 500 spends at least $100,000 annually with Anthropic sparked friendly disagreement. One host found 50% surprisingly low, assuming virtually every Fortune 500 company would trivially spend six figures testing multiple competitive models, including OpenAI, Google Gemini, and open-source models: “The Fortune 500 is probably spending 100K annually on open source models and OpenAI and Google and Microsoft and Anthropic. I would have assumed.” The other host countered by noting enterprise inertia—such as large corporate clients issuing generic boilerplate Gemini instances across internal divisions rather than sourcing multiple bespoke enterprise LLM contracts. Even so, they note cross-pollination is inevitable, pointing out that even researchers and teams at Meta or Google likely spend six figures running Anthropic models internally before the conversation turns toward issues surrounding model distillation.
[10:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=600s) Anthropic’s Distillation Leak, Terms of Service Violations, and the Christmas Revenue Slump
The discussion opens with insights emerging from Anthropic's investor day regarding model distillation—third parties querying Anthropic's Claude models to train and bootstrap competing artificial intelligence architectures. Previous rumors online had wildly exaggerated the scale, with some critics on X asserting that distillation represented half of Anthropic's revenue. However, figures floated at the investor day put illicit distillation between 5% and 7% of Anthropic's estimated $70 billion annual recurring revenue (ARR). The hosts describe these illicit distillation outfits as sophisticated, organized bot networks deploying billions of dollars to systematically extract tokens: “They’re distilleries. They’re illegal distilleries. They’re criminals, basically.”
A critical wrinkle in Anthropic’s S-1 registration statement reveals that this 5% to 7% bucket is officially categorized under general "terms of service violations" rather than distillation alone. While model distillation likely represents the lion's share of high-volume token generation, the classification legally encompasses an assortment of banned activities, including cybercrime, adult content, unauthorized hacking operations, and biological weapons development.
```mermaid
flowchart TD
accTitle: Anthropic Terms of Service Violation Revenue
accDescr: Diagram illustrating how the 5 to 7 percent TOS violation bucket splits across illicit API use cases.
A["Total ARR: ~$70 Billion"] --> B["Legitimate Enterprise API & Users<br/>(93% - 95%)"]
A --> C["Terms of Service Violations<br/>(5% - 7%)"]
subgraph TOS["TOS Violation Breakdown"]
C --> D["Model Distillation Farms<br/>(Bulk Token Theft)"]
C --> E["Cybercrime & Network Hacking"]
C --> F["Banned Content & Bio-Risk Exploration"]
end
```
The conversation then shifts to seasonal API volatility, jokingly noting that the "war on Christmas" deals a tangible blow to the AI sector. According to reported figures, the AI trade suffers a 20% dip during the final two weeks of December. As software engineers, vibe coders, and startup developers go on holiday break, API consumption slows drastically. The hosts joke about Anthropic's chosen stock ticker symbol, noting that adding a dollar sign produces `$ANT`, which humorously resembles the start of Santa. Pondering the ticker name, they banter over Elon Musk’s philosophy that companies eventually become the opposite of their branding:
- Distillation represents an estimated 5% to 7% of Anthropic's roughly $70 billion ARR.
- End-of-year holiday slumps reduce API consumption and developer query volume by approximately 20%.
- Anthropic chose the stock ticker symbol `ANT` ahead of its planned public market debut.
- Speculation over opposites leads the hosts to propose an "anteater" biologically, or an "elephant" or "whale" in terms of market scale, noting that going public around a $2 trillion valuation would make it the largest IPO in history.
---
[13:38](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=818s) Anthropic’s Seven Founders, Class F Golden Share, and 50.1% Voting Control
The hosts analyze reporting from Liz Hoffman on the corporate governance structure detailed in Anthropic’s S-1 filing. Anthropic has structured an unprecedented mechanism to maintain permanent insider governance without requiring matching economic ownership. Prior to completing the offering, Anthropic issues a single share of "Class F common stock"—effectively a "golden share"—to an entity designated as "Founder LLC," whose sole initial members are Anthropic's seven co-founders.
Quoting directly from the filing, the text outlines the rationale and the mechanics: “Our Class F common stock will be entitled to an aggregate number of votes that equals 50.1% of the total voting power of our outstanding shares of capital stock with respect to matters submitted to stockholders generally including the election of our directors other than the class T directors.” The hosts joke that Anthropic missed an obvious branding opportunity by not naming the entity "My Precious LLC," but praise the tactical genius of the LLC umbrella: it creates a unified front in public voting, masking internal splits (such as a 4-to-3 dispute) behind a single monolithic 50.1% ballot.
```mermaid
flowchart LR
accTitle: Anthropic Governance and Voting Architecture
accDescr: Flowchart showing how the Class F single golden share controls 50.1 percent voting power via Founder LLC.
- Anthropic is organized as a Public Benefit Corporation (PBC), pairing its public mission with an LLC voting vehicle.
- The single Class F share holds absolute majority control (50.1%) over ordinary shareholder business, including director elections outside of Class T directors.
- By voting through a single LLC, external observers never see dissenting founder ballots, even when intense internal infighting occurs.
---
[16:30](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=990s) Governance Complexities, Mark Zuckerberg Comparisons, and the $141B Balance Sheet
The conversation digs deeper into the corporate vulnerabilities and edge cases of Anthropic's bespoke governance apparatus. Questions arise regarding the independent board—which includes former Federal Reserve Chairman Ben Bernanke—and whether the 50.1% Founder LLC voting bloc supersedes the board's authority to select and dismiss management. The hosts compare the dynamic to tech industry governance battles, noting that tech journalist Corey Weinberg at *The Information* and *The New York Times* has endless plot twists to cover.
Unlike Mark Zuckerberg at Meta, who must continuously balance share issuance and supervoting dilution, Anthropic’s Class F founders can theoretically liquidate their economic stake entirely while keeping absolute control over the company: “This structure enables them to never, they could all own basically nothing and be happy because they still have the golden vote.” The hosts examine edge cases that could shatter the harmony of the seven co-founders:
- Succession planning: inheritance by multiple children, transfers through divorce, or deaths among the founders.
- Deadlock risk: if a founder dies without heirs and membership reduces to six, the LLC could easily freeze in a 3-versus-3 deadlock.
- Fair market value paradox: though legally stripped of economic rights, the single Class F share wields controlling authority over a multi-trillion-dollar company. Big tech incumbents like Google, Microsoft, or Nvidia would theoretically bid tens of billions of dollars at auction to acquire that single share, raising tax and valuation questions.
The segment concludes with SemiAnalysis’s breakdown of Anthropic’s balance sheet as reported in the S-1:
> “We had raised an aggregate of $141 billion of capital and had cash and cash equivalents of $112.4 billion and short-term investments of $19.3 billion.”
With $131.7 billion in liquid reserves remaining out of $141 billion raised as of June 30, 2026, Anthropic had burned a total of only $9.8 billion across its lifetime while establishing historic top-line growth.
[20:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1200s) Anthropic's $2T IPO Valuation Debate and Spark Capital's 131x Bet
Anthropic has turned profitable on an operating income basis as of the June quarter, prompting the hosts to marvel at its unprecedented trajectory. As John Coogan remarks: “This is likely the most efficient scaling of a business in the history of capitalism.” The upcoming IPO brings massive speculation: while the ticker symbol and its playful reference to Santa Claus are known, the exact amount raised remains rumored at around $100 billion. The expected valuation is pegged between $1.5 trillion and $2 trillion, with $2 trillion very much on the table. Even with 5% to 7% dilution, the listing will put billions on Anthropic's balance sheet, marking a staggering up-round. The IPO is expected the week of November 16th, aligning with TBPN’s live broadcast from Tomorrow X, where they plan to host major capital allocators to debate the offering live.
```mermaid
flowchart LR
accTitle: Spark Capital Investment Progression
accDescr: Diagram tracing Spark Capital leading Series C through Series H to a projected 18 billion dollar stake.
subgraph Spark["Spark Capital Investment Bet"]
C["Led Series C<br/>(131x Return)"] --> D_H["Follow-ons:<br/>Series D through H"]
D_H --> Total["Total Invested:<br/>$777 Million"]
end
Total --> Stake["~$18 Billion Stake<br/>(@ $2T IPO Valuation)"]
```
The S-1 filing contains unconventional risk factors, including catastrophic AI risk and existential "doom" scenarios, but the market has largely shrugged them off. Investors cannot realistically model existential doom into a discounted cash flow (DCF), taking the stance that there is little to adjust other than hoping it doesn't happen. Curiously, while marketing spend has exploded—from a mere $2 million in 2023 to $268 million, and up to $3.17 billion in the first half of 2026—the hosts notice no dedicated cost line item for AI safety. For a company putting catastrophic risks directly into its regulatory filings, one would expect AI safety compute and headcount to scale visibly as distinct expense lines alongside revenue.
- **Advertising Stance:** Despite holding an anti-ads stance regarding their own product business model, Anthropic has become the single largest advertiser in the AI sector, spending billions.
- **Hardware Commitments:** The S-1 reveals a $125 billion equipment lease arrangement with Broadcom.
- **Cloud Marketplace Revenue:** Approximately 48% of Anthropic's revenue flows through cloud marketplaces, sparking debates about revenue recognition that operating income figures ultimately settle.
- **Venture Windfall:** Spark Capital's Yasmin Razavi led their Series C round (yielding a 131x return) and doubled down through Series D through H, committing a cumulative $777 million. At a $2 trillion valuation, that stake is worth roughly $18 billion, cementing her position for the top of the *Forbes* Midas List.
---
[25:12](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1512s) Anthropic's White-Hat AI Hack and the $20B Bug Bounty Debate
Right as the S-1 dropped, news broke regarding security vulnerabilities inside Anthropic. Reporting in *The Washington Post* revealed that a team of ex-DOGE cyber researchers penetrated Anthropic's internal systems—including Slack channels discussing their most sensitive models—in under 12 hours using AI agents. Conducted under a white-hat safe harbor, the breach was seemingly timed by Cathedral to go public on the day of the IPO to drop a bombshell. However, the sheer strength of the S-1 financial figures completely overshadowed the security incident.
The hosts explore whether the federal government should step in to finance national cybersecurity initiatives on the scale of an industrial bill. A major critical infrastructure failure—such as a power grid failure or banking system collapse—could trigger an immediate, massive shock to GDP akin to the COVID-19 pandemic. Right now, frontier labs like OpenAI and Anthropic run private vulnerability initiatives (such as Daybreak, Mythos, and Glasswing) partly for defensive posture, and partly for the public relations victory of disclosing and patching bugs.
```mermaid
flowchart TD
accTitle: Government Bug Bounty Concept Flow
accDescr: Flowchart illustrating a hypothetical 20 billion dollar federal bug bounty and its potential consequences.
Coogan and Hays debate what would happen if the federal government introduced an official $20 billion bug bounty program to secure American infrastructure. While a huge government payout would motivate frontier labs, it could also trigger absolute chaos: online course sellers in Miami hawking tutorials on how to point "obliterated" Hugging Face models at critical targets to chase six-figure government bounties. As one host observes when weighing the hacker incentives: “I think you'd be surprised by the number of hackers who would break into something like Quest Labs and be like, between having a bunch of Bryan Johnson's blood work and just a check from the government for six figures, I'm taking the check for the government.” Still, the risks of incentivizing untrained actors with zero experience handling sensitive records remain severe unless program participation is tightly restricted to vetted institutions.
---
[28:57](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1737s) Samsung Smart TV Spying Allegations and the Case for Dumb TVs
The discussion pivots to consumer hardware grievances, sparked by an account named Log4j detailing intrusive tracking behavior on Samsung Smart TVs. The hosts vent about the modern state of displays, expressing a deep desire for a premium "dumb TV" in 2026. Rather than an ad-laden, lagging operating system, consumers just want high-quality display panels that turn on instantly, handle color correction, match input frame rates, and display external inputs cleanly without forced software updates.
Today's smart TV market offers virtually no high-end displays stripped of software bloat because manufacturers subsidize display hardware by turning panels into ongoing surveillance and advertising platforms. As one host laments: “Everything is ad-supported. Everything is spyware and calling home and listening to you. And it's all very annoying.” The problem goes beyond privacy violations—the bloat actively ruins the hardware experience, making simple navigation sluggish and buggy. While cheap streaming devices from companies like Amazon also push ads, the hosts argue that paying a premium price for a display should free the user from spyware and slow interfaces entirely.
[30:00](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1800s) Samsung Smart TV Spying Allegations and Privacy Scrutiny
The hosts break down reports from a developer hacking a Samsung Smart TV who discovered that the screen is effectively watching the viewer right back. Deep in the code, researchers found that the television loads a region-specific convolutional neural network designed to track app usage granularly. For example, when running an app like YouTube, the system actively monitors which specific videos you watch, which advertisements you choose to skip, and your exact watch-time progress through a piece of content. The hosts also point out that code for facial recognition was uncovered. While there is a brief debate over whether the facial recognition is analyzing faces on screen—identifying if someone is watching MrBeast—the hosts argue it is far more likely analyzing who is sitting in the room watching the display.
The discussion turns to the double standard in tech regulation and consumer privacy. Whereas device manufacturers like Apple face intense scrutiny and build elaborate privacy controls, and Meta faces pushback to the point of installing physical recording indicator lights on its camera smart glasses, the television hardware industry manages to operate almost entirely under the radar. Because television manufacturers do not face the same regulatory pressure or public outcry, they are able to aggressively monitor users' living rooms without clear oversight or accountability.
> “And the TV industry just isn't under the same amount of pressure or scrutiny. So I think they can just... kind of do whatever they want.”
- A developer digging into Samsung Smart TV software uncovered a region-specific convolutional neural network (CNN) tracking app interactions.
- The tracking monitors YouTube video choices, ad-skips, and playback progress.
- Code references for facial recognition indicate potential tracking of the viewer's identity in the living room rather than just on-screen content.
- Legacy hardware manufacturers bypass privacy standards that Silicon Valley giants like Apple and Meta are forced to navigate.
---
[31:15](https://www.youtube.com/watch?v=Hw5YREMzgJ0&t=1875s) Instagram Takes on YouTube With Long-Form Video
Instagram has officially announced tests for long-form horizontal video formats referred to as "Features," positioning itself as a direct competitor to YouTube. The platform's goal is to become an all-in-one entertainment surface where users transition seamlessly from watching a 90-second short on their phone to streaming a 90-minute piece of content directly on their smart TV. Instagram is simultaneously experimenting with live broadcasts embedded directly inside the Reels tab alongside a dedicated hub for episodic series. The hosts note that Instagram head Adam Mosseri discussed this broader video expansion strategy during an appearance on Colin and Samir's show. Recalling historical battles over platform exclusivity, the hosts cite how YouTube previously threatened to destroy creators' reach when Netflix attempted to recruit top YouTube personalities.
The hosts explore the structural friction of modern online video: viral short clips frequently strip context and spark outrage without converting casual viewers into long-form subscribers. In their own experience running TBPN, the core daily live shows and daily "Diet" recap cuts grow steadily day over day, but isolated viral clips rarely translate into subscriber spikes, with conversion rates hovering around "two basis points." Viewers often take short clips out of context—such as omitting the original question asked—and direct outrage at statements that were clarified mere seconds outside the clip's boundary. If Instagram allows a viewer to tap a viral clip and seamlessly jump to the exact timestamp in the full horizontal video without leaving the app, the hosts believe it will fundamentally improve the health of online discourse by providing immediate access to the entire conversation.
> “Right now, we have a big problem. And we may have played a part in it, but the big problem is that you get a clip, right? And you get this really interesting moment... And some percentage of the people watch the clip and then go find the full length thing. But most of the people get like they feel some emotion, whether it's joy or rage.”
```mermaid
flowchart LR
accTitle: Instagram Clip to Full Video Transition
accDescr: Diagram showing how Instagram's proposed Features product connects short viral clips directly to timestamped long-form horizontal videos.
subgraph ShortForm["Short-Form Discovery"]
Reels["Reels / Clip Feed"]
Snippet["30-90s Viral Highlight"]
Reels --> Snippet
end
subgraph Conversion["Seamless In-App Handoff"]
OneTap["One-Tap Timestamp Link"]
Snippet --> OneTap
end
subgraph LongForm["Horizontal Feature"]
FullVideo["Long-Form Feature (Up to 90 mins)"]
Context["Context Restored (Reduces Outrage)"]
OneTap --> FullVideo
FullVideo --> Context
end
```
- Instagram is piloting horizontal long-form video ("Features"), live broadcasts in Reels, and a dedicated series hub.
- The initiative aims to capture both mobile short-form audiences and living room smart TV viewership.
- Conversion from viral short-form clips to long-form episodes is estimated by the hosts at just a fraction of a percent (~0.02% or two basis points).
- In-app linking directly from short excerpts to the exact timestamp of the full video helps eliminate manufactured outrage caused by un-contextualized clips.
## Mentioned
Adam Mosseri, Amazon Web Services, Anthropic, Apple, Ayal Harry, Ben Bernanke, Bloomberg, Broadcom, Bryan Johnson, Cathedral, ChatGPT, Claude, Colin and Samir, Corey Weinberg, Ed Zitron, Forbes, Google Cloud, Google Gemini, Hugging Face, Instagram, Jeff Bezos, John Coogan, Jordi Hays, Lambda, Liz Hoffman, Log4j, Mark Zuckerberg, Meta, Microsoft Azure, MrBeast, Netflix, Peter Thiel, Quest Labs, Samsung, SemiAnalysis, SpaceX AI, Spark Capital, TBPN, The Information, The New York Times, The Wall Street Journal, The Washington Post, Yasmin Razavi, YouTube
## Video description
Sign up for TBPN’s daily newsletter at TBPN.com
Diet TBPN delivers the best of today’s TBPN episode in 30 minutes. TBPN is a live tech talk show hosted by John Coogan and Jordi Hays, streaming weekdays 11–2 PT on X and YouTube, with each episode posted to podcast platforms right after.
Described by The New York Times as “Silicon Valley’s newest obsession,” the show has recently featured Mark Zuckerberg, Sam Altman, Mark Cuban, and Satya Nadella.
TBPN is made possible by:
Ramp - https://ramp.com
Public - https://public.com
Cisco - https://www.cisco.com
Console - https://www.console.com
CrowdStrike - https://www.crowdstrike.com
Figma - https://www.figma.com
MongoDB - https://www.mongodb.com
NYSE - https://www.nyse.com
Railway - https://railway.com
Shopify - https://www.shopify.com/
Anthropic's S-1 registration statement reveals historic top-line growth and positive operating margins achieved through a pure compute-rental model, alongside an unprecedented corporate governance structure that grants its seven founders absolute voting control without corresponding economic equity.
Financial Scale and the Compute-Rental Model (4:08)
According to disclosures in its S-1 filing, Anthropic reported an annualized revenue run rate of $67.2 billion—growing 45% quarter-over-quarter—backed by $131.7 billion in liquid reserves and lifetime capital burn of $9.8 billion out of $141 billion raised. Operating margins turned positive at 19%, while compute costs as a portion of cost of goods sold fell from 72% to 48%. Rather than financing physical data centers, the company spent just $482 million on first-half capital expenditures, relying instead on a multi-year cloud commitment backlog totaling roughly $637 billion across Google Cloud ($114.5 billion), Amazon Web Services ($110 billion), Microsoft Azure ($45 billion), Broadcom ($161 billion), and SpaceX AI (up to $120 billion).
Anthropic disclosed over 1,000 customers spending at least $1 million annually and more than 10,000 spending over $100,000, yielding a 900% net revenue retention rate across those tiers. While 50% of the Fortune 500 spends at least $100,000 annually on Anthropic models, program hosts John Coogan and Jordi Hays noted that enterprise inertia and default deployments of rival tools like Google Gemini mean large corporations often maintain fragmented adoption rather than standardized enterprise-wide contracts.
Class F Golden Share and Founder Governance (13:38)
Anthropic structured its public debut as a Public Benefit Corporation using an unusual voting mechanism to ensure permanent insider control. The company issued a single share of 'Class F common stock' carrying zero economic interest to Founder LLC, an entity governed exclusively by Anthropic's seven co-founders. This single golden share exercises 50.1% of the total shareholder voting power over ordinary corporate matters and the election of directors outside Class T seats.
This architecture permits the co-founders to divest their economic ownership entirely while retaining voting supremacy, masking internal founder disagreements behind a single institutional ballot. However, the hosts emphasize structural governance vulnerabilities inherent to the setup, including succession complications from founder deaths or divorces, potential 3-versus-3 deadlocks if membership shrinks, and the paradox of valuing a non-economic share that wields absolute authority over a multi-trillion-dollar firm.
Valuation, Distillation Risks, and Cybersecurity (20:00)
Ahead of an anticipated mid-November IPO targeting a $1.5 trillion to $2 trillion valuation, the filing details operational risks, including seasonal API usage declines of 20% in late December and illicit model distillation. Anthropic revealed that roughly 5% to 7% of its recurring revenue is linked to terms-of-service violations, an umbrella covering unauthorized bot distilleries querying Claude to train competing models alongside cybercrime and prohibited use cases. Early backer Spark Capital stands to convert a cumulative $777 million investment across Series C through H into an estimated $18 billion stake at the top-end valuation.
The S-1 disclosure coincided with a report that ex-DOGE cybersecurity researchers penetrated Anthropic's internal Slack infrastructure in under 12 hours via AI agents. While the security incident was overshadowed by the company's financial figures, the breach prompted debate between the hosts over whether a multi-billion-dollar federal bug bounty program could harden critical national infrastructure without inadvertently incentivizing rogue, unqualified actors to target sensitive systems.
Consumer Surveillance and Platform Video Shifts (30:00)
Beyond enterprise AI, the hosts addressed privacy concerns in consumer hardware after a developer uncovered an active, region-specific convolutional neural network and facial recognition code embedded in Samsung Smart TVs. The software tracks granular viewer behavior within third-party applications like YouTube—including specific videos watched, ad-skips, and watch-time progress—underscoring how television manufacturers subsidize hardware margins through ambient tracking and programmatic advertising while facing far less regulatory scrutiny than smartphone and social platform operators.
Concurrently, Instagram launched tests for horizontal long-form video dubbed 'Features,' adding dedicated hubs for episodic series and Reels live streams to compete with YouTube across mobile devices and connected smart TVs. The hosts argue that bridging viral short clips directly to full-length videos at exact timestamps addresses a core structural flaw of online video: low conversion rates between short clips and full episodes, alongside the tendency of isolated snippets to trigger outrage by detaching arguments from their broader context.
Anthropic's pending public debut presents public markets with a profitable, highly capital-efficient hyper-growth model, yet leaves open questions regarding whether its monolithic founder voting control, multi-hundred-billion-dollar hyperscaler commitments, and reliance on rented infrastructure will withstand competitive pressures and long-term public scrutiny.
[0:00] - This is Anthropic IPO. It's the most important story of the day, after the amount of days till Christmas. Let's count it down. We're very close. 59 days till Christmas. Do you think Anthropics is 59 years old? There you go. Yeah, it's interesting. Anthropics seems to be sending us incredible innovation. Yes. at tbpn that we tie random fun facts yes to the christmas countdown i i also like that just to clarify it's not that peter teal is 59 years old today right No, he's just 59. I think his birthday was two weeks ago. I think his birthday was two weeks ago. He just happens to be 59. It just happens to be that age. So it's just a fun fact. Because over the weekend I was thinking about we did Jeff Bezos' 62 on the 62 days till Christmas and I was like, was it Jeff Bezos' birthday or was that just a random fun fact? And it appears to be just a random fun fact. Anyway, so do you think Anthropic by dropping the S1 on exactly 59 days till Christmas is trying to send like some sort of subliminal message? What would that message be? Oh, just that it's 59 days, so you should go into the S1, control F for the number 59 to find the hidden meaning of the S1. Or go to page 59. Go to page 59. Maybe you should. Have you been yet? Oh, I've been to page 59. I think I know exactly what they're trying to say with this. It's 59 days till Christmas. Open up the S1, control F, hit 59, page down to page 59. What do you see? It says, we may not have the ability to raise the funds necessary to repay the principal and accrued interest of our outstanding. Yeah, it's just a random page. I don't think there's anything going on. Anyway, there are a bunch of interesting facts. You've got to dig through it. The S1 is huge, obviously. I mean, it's a massive company raising a massive amount of money. has not been listed yet. There's sort of speculation in the Wall Street Journal around $100 billion raised. Bloomberg contextualized the IPO filing as, Anthropic files for IPO in global race to raise cash for AI. The interesting thing is that Anthropic has a lot of cash. They also have a lot of revenue, 67.2 billion in annualized revenue. That's up 45% quarter on quarter at this scale. Incredible. Super, super, super fast growth. And full Q3 numbers aren't in the initial S1, but that number that was rumored from the Anthropic Investor Day is now in the actual S1. Cash on hand, $131.7 billion. That's a lot of cash, and they've been super capital efficient. People are saying it might be the most capital efficient company in history. They, of course, brought in $110.8 billion in the first half. period. And the S1 came out, I think in some ways, like the leak around the S1 that had happened maybe a week or two ago around this like... And the number that everyone fixated on was this $80 billion loss, this accounting charge. I think people now have had an opportunity to dig in here. And the funny thing is the nature of having access to Claude and ChatGPT and other agents is people can process the S1 a lot faster. So we're already getting like a lot of, it used to be like an S1 release and the VCs would have 12 hours. Well, they've had 12 hours to pull out interesting stuff. Now it's like it's all coming out in real time. And it is totally remarkable. I think it's amazing for the AI trade. overall. It feels like a beat. It feels like this beat expectations. Totally. It didn't like surprise, oh, they're actually doing 100 billion in revenue or something. But I mean, really, really solid people haven't found anything that's like, oh, I didn't understand that the business was actually shaped a very different way or something like this. This sort of meets expectations. I yield Harry said, you know, it's a decent S1 when Ed Zittron hasn't tweeted anything. Yeah, I think this is a beat on the expectations of the S1. Command F, he has to command F, right? Command F, bad. Command F, sad. Command F, problem. The NASDAQ is up 1%, and it feels like the AI trade is starting to get put back on. Obviously, there's been a lot of selling in the further risk on, the further out on the risk curve assets in the AI trade. Some of the neoclods down a lot. clouds like there's this interesting thing where they raised 110 billion dollars they have 130 billion dollars in cash how much did they spend in the first half of the year on capex this is the company that wants to serve ai at scale and how much do they actually spend buying computers building computers that they own 482 million dollars less than half a billion dollars.
[5:00] Now, of course, over time they'll build more, they'll do all sorts of stuff. Yeah, they've been renting computers. At least in the first half of the year, the strategy has been to rent. And then also they have a pretty huge amount of backlog with computers. the hyperscalers. It looks like this scary number, I think it's like $690 billion of commitments. Really not that much, 10x of revenue currently if they keep scaling. But over a multi-year period. Yeah, yeah. So there's Google, Google, Amazon and Microsoft are all in the 40 to 130 billion dollar range. I can pull up the exact numbers, but the but the contracts are how much that they have to spend on compute through 2033. Some of them go all the way out to 20. which is really, really, really far. So they have a lot of time to actually work through that. So Google Cloud, their commitment is $114.5 billion, and that's minimum cloud spending through July of 2033. So they're just renting, and it's over a long period of time. Is there any scenario where they don't have the demand to actually parcel out $20 billion of spend over the next six years, seven years? It seems like they're in a good position there. AWS is at $110 billion, and that's stretched out to 2036. And then Azure is only $45 billion, and that's minimum cloud spending through 2033. Now, they do have $120 billion with SpaceX AI. That's up to, that's the maximum. is at 45, LAMD is at 42 billion, and Broadcom is at 161 billion for a total of 637 billion in spending commitments, but again, over five, six, six, six, six, six, six, six, six, six, maybe eight years in some of these scenarios. So pretty healthy there. Um, The other interesting surprises, I mean, people are pretty happy about the margins. So compute cost as a function of revenue, they were at 72%. You can think about this as like cost of goods. 72% of every dollar they made was getting eaten up by compute costs. That's down to 48% compute costs divided by revenue. So that's rapidly improving efficiency quarter on quarter. They went from negative 13% to 19%. operating margin. And then the big number that people are excited about is the big customers. So they have, I think, over a thousand customers spending over a million dollars, something like that. And then over 10,000 spending over 100K over. Yeah, over 10,000 spending 100K and net revenue retention is 900 percent on those. So they got in early in 2024, 2025, and then all of those using the APIs here and there. And then we went to the token maxing era. They were in the right place. And so many companies scaled their spend that on average, it was basically a 10x increase for organizations. But that makes sense. We've all seen this as, you know, more and more people are like, hey, it's not just like the early adopter on the team that has an AI agent running all the time. It's actually spread to everyone. It's spread to the marketing organization, the operations teams using it. And so these services are going out all over the place. half of the Fortune 500 spending over $100,000 annually. that actually felt low to me. Obviously it's a really impressive, like it's great, but I was under the assumption that all of the Fortune 500 was spending a lot on every single model. And I would have said, oh, of the Fortune 500, 100K annually. The Fortune 500 is probably spending 100K annually on open source models and OpenAI and Google and Microsoft and Anthropic. I would have assumed. You've got to ask what that half is, what they're up to. I mean, some of them might just be in industries that don't write any software, really. But it feels like every company, if you're in the Fortune 500, you have some software. It's kind of bearish any way you put it. I do know that this was maybe six months ago, but I was talking to someone. at a Fortune, relatively high up at a Fortune 500, and they said that they had, they were basically running like, boilerplate Gemini that the whole org had access to, and that was their AI tool. They were telling everyone, "Just use this." So there's, I'm sure, examples of that. Yeah, yeah, yeah. And so yeah, maybe you see like 50% has spending over 100K on Anthropic and then there's some slice that's on OpenAI, some slice on Gemini and Google, and then some slice on some other section of the market. But I mean, it's odd because like, even meta is probably spending over over 100k annually on anthropic i have to imagine that that google employees some of them have access to anthropic models to the tune of 100k and so that was sort of interesting uh the other interesting update was on the distillation
[10:00] So coming out of the investor day, it felt like we finally got a number on how much of this revenue is distillation, how much of it might go away if Anthropic figures out how to prevent distillation. And the number that was floated was five to seven percent. Totally reasonable. It was not like a it's over moment. It was like, oh, like that's significant. Thank you. 5% of $70 billion ARR. It was way less than some of the peanut gallery on X would like to say. Half their revenue. That's ridiculous. But still, it's crazy that there are companies out there, or I don't even know what you call them, bot farms, organizations. Distilleries. They're distilleries. They're illegal distilleries. They're criminals, basically. They're TOS violation companies. But they're networks. They're organized. And what's interesting is that it's not just, you know, oh, some, you know, enterprising, scrappy data broker sales guy, like, figured out how to, like, puppeteer a couple clawed accounts and sneak some tokens out and, like, resell them into, like, somebody who's selling it to Kimmy or whatever. It's like, no, people are deploying, like, billions of dollars to this strategy, and it seems pretty organized. At the same time, the wrinkle that we got from the S1 is that that 5% to 7% number that was leaked out from the investor day, it's not explicitly listed as distillation revenue. It's just terms of service violation. So it could be people that are doing adult content or cybercrime or hacking or bio weapon development anything that's going to get you banned. I would imagine that the vast majority of it is still distillation because what else is going to generate that many tokens that not much revenue, but it is worth noting that it's not entirely distillation attack. I like that we still have the Christmas trim up. It's very, very good. The other one, speaking of Christmas. There apparently is a war on Christmas going on, and it's hurting Anthropics significantly. We have some real numbers on how much Christmas hurts the AI trade. It's apparently 20%. So API consumption, respectively revenue in December for the last two weeks will be down 20% as engineers, as vibe coders, as startups, as everyone goes on holiday. Okay. They're celebrating Christmas. People will be buying. Well, there's been some speculation. Obviously, they announced what their ticker is going to be, ANT. during the road to Christmas. But if you put the dollar sign, it looks like the start of Santa. And so speculation that they're sort of, yeah, it's kind of a nod to Santa. Yeah, yeah, yeah. You know, that does make perfect sense. What is the opposite OF AN ANT. Like the animal. opposite of an ANT. Like if you had to think of like the animal, the ant, and then the opposite. Because this is the whole Elon Musk philosophy is like you're destined to become the opposite of whatever you are branded as. So the biological opposite of an ant It's an anteater. So maybe they're destined to become an anteater. I don't know. Size opposite would be an elephant. It is an elephant of a stock. Going out of $2 trillion is going to be the biggest IPO ever. Maybe that's the destination. I was thinking whale. Well, Yeah, I mean, it's good for all-- there can be a lot of whales moving into the stock, right? Anyway, Liz Hoffman. It says seven people. So basically, she's pulling out the founder voting structure in the S1, which spells out, we believe that our seven co-founders. as a group are distinctly equipped to be stewards of our mission. Immediately prior to the completion of this offering, we will issue a single share. They're calling it the golden share of Class F common stock. They're not actually calling it the golden share, but it is a golden ticket. The Class F common stock, a special class of our common stock to a limited liability company. of which our seven co-founders will initially be the sole members of the "Founder LLC." Our Class F common stock will be entitled to an aggregate number of votes that equals 50.1% of the total voting power. of our outstanding shares of capital stock with respect to matters submitted to stockholders generally including the election of our directors other than the class t directors so um missed opportunity to call it uh my precious llc but i think that uh founder llc if they actually have founder llc that's very cool because it means like they some obviously like somebody over the years would have sniped that LLC, and then they could either sort of grant the sort of grant the name to someone else. They can sign off on it. It might just be a placeholder name. But it seems like they might actually have it, which is very cool. And yeah, it's kind of funny that we
[15:00] Like, OK, we have this public benefit corporation And that's super important. for... for economic redistribution. Yeah, and then you have just a good old-fashioned LLC that can offer one vote. Yeah. The thing that would have been interesting is like if they were if this would have been seven votes and that they would have been public, like where the line was, but this allows them to effectively have their own disagreements internally, not have there any be any visibility externally. Interesting. Okay, so that's the best. So you could have like, you could have like, it'd be like a four, three vote. Yeah, people are like, you know, massively infighting, but all all we see is a single vote. Yeah, got it. Okay, interesting. I thought it was more about Yeah. Would your like if you die and you in your will, you're like, oh, I want to like pass this LLC down to my five children. Then do they become voting members? Do they have a voting a separate? I mean, the LLC probably has its own voting structure, right? Yeah. So it could be within that LLC – There could be one member of the LLC that has 90% of the shares in that LLC, so they could have one person control. potentially. Like we have no idea the structure of that LLC. which is interesting, right? Yeah, we just know that we know it's... they're precious. What's also interesting is like I don't understand how the board works now because there's the there's the board that's independent with like Ben Bernanke on it and they appoint the executives, I think, but they were supposed to have control. But now it seems like this has control that supersedes that. So there's a lot of interesting uh... There's a lot of interesting complexity. Yeah, it's really full employment for Corey Weinberg. I understand why the New York Times was like, we've got to get this guy from the information because he's going to have full-time employment. Because – Every time he writes a breakdown or a scoop or something, Anthropic's like, here's another twist. Here's another plot twist. Good luck figuring this one out. It's a very, very interesting story. Set up. I wonder if this will become more common. Like with AI. Well, yeah, so Zuck is in this position where later on, like he's had like total control of meta. Now he's like, "Okay, I need to raise equity." And so he's creating this new class so that he doesn't dilute himself so much. In the fullness of time, he could be diluted down. This structure enables them to never, they could all own basically nothing and be happy because they still have the golden vote. The golden vote. The golden vote. Interesting. I'm so interested to understand how this LLC votes because, I mean, even if it's just seven equal founder votes, which is cool. Like you got seven co-founders, everyone's on the same page. That's awesome. You'll never have a three, three V three, but like what happens in succession planning? Like what happens if someone has multiple kids? What if somebody gets divorced? What if somebody dies? Like there are these weird things that like the perfect structure probably doesn't hold up under. Like if one of the co-founders dies with no children and they're just like, I don't – then it's just six to six and it's six people. Then you're back to 3v3. Is that a problem? I don't know. Like there's just so many different dynamics of what – of what could happen? We need to study the dynamics of ant colonies. Maybe. I don't know. Maybe that's what it is. Or maybe it's elephants. I don't know. Yeah, our class F common stock will not have any economic rights. Interesting. Class F common stock entitled to 50.1% of total voting power. I wonder how that's treated for tax purposes. They say it doesn't have any economic rights, but if it has incredible power over something valuable, can it have a fair market valuation? Certainly not. There are plenty of people that would pay a lot for that one share, right? If you were to market to market and you just ran an auction, Like, you know, like Google and Nvidia and Microsoft would all be like, well, yeah, I'll pay. like a billion dollars for that share, or 10 billion, or 100 billion dollars for that share, because I would get control over this amazing asset. So like, it has value, but you just never market to market? I mean, it could be written into the operating agreement that that They can't be transferred or sold. You could do a forward contract against it. The most mind-blowing stats Semi-Analysis has seen about the Anthropic S1 is as of June 30, 2026, "We had raised an aggregate of $141 billion of capital and had cash and cash equivalents of $112.4 billion and short-term investments of $19.3 billion." It means today they've burned $9.8 billion. They're growing at rapid pace in the next year.
[20:00] profitable on an operating income basis as of June quarter. This is likely the most efficient scaling of a business in the history of capitalism. So when is this going out? What do we know? What don't we know? We know the ticker dollar sign and obviously reference to Santa Claus. Yes. But we don't know how much they're raising. It's rumored to be a hundred billion. We don't know the valuation. People are saying somewhere between 1.5 or 2 trillion. 2 trillion feels like it's on the table now. When the market was selling off and there was like a lot of discussion over the Probably still could. It doesn't really matter at the end of the day. It's still a huge up round based on the last. evaluation puts a bunch of money on the balance sheet whether it's five percent dilution seven percent dilution not the biggest deal the ipo should be happening the week of november 16th as a reminder we will be at tomorrow x around that time and It should be an incredible time for us to be doing the show because we have an insane guest lineup with some of the biggest, greatest capital allocators in tech and business. And so we will be able to get everyone's take live as this is progressing. Whether even if it's delayed another week, we're going to have a lot of great stuff to talk about as everyone gets ready to position. Maybe go along, give their thesis, give their take. Maybe they're already in Anthropic. Maybe they bought pre-IPO shares. Risk factors. Sharon, a satirical risk factor. What does he actually say? People are talking about time travel and stuff. What else is in there? The only one, I mean, everyone sort of knew that the IPO risk factor was going to have catastrophic risk, sort of like the AI doom, the doom scenario. It did make it into the S1. The market seems to have digested it. And been like, yeah, we know. Yeah. You guys have been talking about this for a long time. we're still going to underwrite it Like there's not really anything you can change about your underwriting. You're not going to put that in your DCF. You're just going to say, OK, I hope that doesn't happen. What's interesting is it's not even a cost Line I'm like we've seen the marketing dollar scaled significantly. So in 2023 they spent just two million dollars I'd love to know what they were spending two million bucks on in 2023 for ads Was that like was that like Facebook ads or something like what were they advertising? I don't remember there any had been a lot ads in 2023 because they were sort of like They just couldn't you couldn't you bundle like events? Maybe maybe okay. Yeah, yeah, yeah, maybe they did like a dev day or something $268 million and then first half 2026 they're they're spending 3.17 billion dollars on marketing lots of money in marketing I didn't see a line on safety I didn't see like a blind item on like how much are they spending on AI safety which is sort of interesting I don't know if I just missed it but if you're putting the catastrophic scenario in the risk factors you would think that that would also flow to other parts of the income statement as a cost item because you would say hey look like One of the weird things about this business is that we invest a lot of compute, a lot of cost, a lot of headcount in AI safety. And so expect that to grow with revenue or expect that to grow with headcount or expect that to grow, whatever. They're doing great. You mentioned, what, $6 million per employee revenue scale or something like that? Yeah, somewhere around there. It's interesting how many billions of very, very anti-ads as a business model but a lover of ads as a business. Oh, interesting. Billions. makes them the number one advertiser, certainly the number one advertiser in AI. I would be surprised if Google could have topped that, maybe. Who knows? No, no, no. That is a very, very interesting... AND I THINK THAT'S A Scenario. Anyway, what else? They have a $125 billion equipment lease with Broadcom. 48% of revenue is coming through the cloud marketplaces, so that was something people were talking about, how much is accounted for in that different revenue recognition way. It doesn't really matter when you look at the operating income. Clearly, that factors that out, but that was a question that people had, and then AI accelerating beyond its limits to... Spark Capital, Vivec pulled out the amount that they had invested at different points, or at least an estimate of that. And you can see they started by leading the Series C, which they have 131 X on, and then they just kept doubling down through D through H, Total of $777 million invested, which would give them roughly an $18 billion stake at a $2 trillion dollar. market cap. So really fantastic. return, early conviction, and Yasmeen, is the only venture capitalist on the board. Yeah. So, yeah.
[25:00] You know what I'm thinking? What? Top of the Midas list. Oh, for sure. Yeah. Like she should probably just jump straight to the front. Yeah, hard to make an argument against that. Okay, so... So they got hacked. Yeah, so Cathedral... You know what they were trying to do. This was a white hat safe harbor. Yeah. This was not the same thing that you've been seeing. So Elizabeth in the Washington Post says, in less than 12 hours, a group of ex-doge cyber researchers were able to penetrate anthropic systems using AI agents, including Slack channels, where the most sensitive models are discussed. This is just weeks after anthropic whistleblowers. It looks kind of silly. There we go. There we go. Joseph Coxon charged the company with launching a hubristic gamble from a private company's slack. Launching a new brisk? Anyway, so the interesting thing was Cathedral seemingly wanted to drop this the day of the IPO. Yeah. Assuming that it would just be this like crazy bombshell. Yeah. And I do think it's an important, interesting story. Yeah. But the S1 was just so good that... It's kind of drowning it out. We were talking about this over the weekend. You were saying you think that there should be like... at least a 10, 20 billion white hat program potentially – I don't fully understand the scale of the economic damage of, like, the Quest Labs hack and the other, like, how serious is the risk that, like, the banking infrastructure shuts down or, like, the grid goes down? But, like, if there's really a risk of, like, the grid going down, you're looking at economic damages in the billions, if not hundreds of billions. Like, you'd just be so disruptive. sudden you just have like a 20% nuke in GDP, like we did with COVID, where everyone was like, we can't go out to restaurants. So the restaurant industry is just zeroed for this quarter. And if something like that happens, like the economic damage is so high, it's You have to incentivize it. Right now there is an incentive which is just Anthropic and OpenAI are like racing with each other so they want to be like the savior. And so they have Daybreak and Mythos or Glasswing to to like, you know, kind of help out and make money and also earn PR wins for disclosing bugs and patching things. But it would be interesting to see it go a lot further. I was copying it to like, what was the Intel deal? Like 20 billion or something? Like if the US government was just like, we're doing a $20 billion bug bounty, like let's secure America. And all the labs are like, okay, well, like this is actually going to move the needle for us. People in Miami are selling courses on how to get your first white hat bounty. Go to hugging face, get an obliterated model. Yep. Go point it. I mean, but that's a pretty good use of time and money. Like you need a lot of people to, to test a bunch of different ideas, all the long tail stuff, figure out, uh, it would be very chaotic. It'd be extremely chaotic, but if it's all white hat and they're all doing it because like, yeah, I got it. How white hat can it be if there's a guy in Miami selling a course on how to use obliterated models to hack I think you'd be surprised by the number of hackers who would break into something like Quest Labs and be like, between having a bunch of Brian Johnson's blood work and just a check from the government for six figures, I'm taking the check for the government. I don't actually care about the underlying data. It would be just extremely chaotic and potentially have some massive negative externalities. You're incentivizing a bunch of people that have never... that have no experience handling sensitive data. Yeah, but you could also restrict the bug bounty program just to big companies. An account called Log4j has been having an absolute field day with their Samsung TV. over the last few days. Where should we even start? I mean, at the high level, I love the idea of getting a dumb TV in 2026. The smart TVs, we've complained about them a ton. Palmer Luckey's complained about them. Like, they're slow. You always have to update them. They have a whole bunch of unnecessary features. You just want it to turn on and look good. You probably need a settings panel. with some of like the color correction and some of the, some of the, some of the decisions that you make to make sure like the frame rate matches, whatever your input sources. But by and large, I don't need the ads, but I also don't need the ability to stream Netflix on the device directly. I just want it to be a dumb panel and that's, Just a product that's not for sale. Right now. Everything is ad-supported. Everything is spyware and calling home and listening to you. And it's all very annoying. Even if you're fine being spied on, it's just annoying because it's slow. It's just a bad product. Amazon has a ton of ads, at least at low cost.
[30:00] Exactly. Yeah. Right. But and so Log4j has been hacking his his Samsung TV and seeing some solid progress, says another development might. Realizing that the TV is watching. He's watching the TV while the TV watches. It's brutal. So the TV has a camera. Apparently the Samsung TV loads a region-specific convolutional neural network, which tracks your usage in the app. Like in YouTube, it's looking for videos you watch, ads you skip, progress in a video. So glad they're interested in me. Code for facial recognition within the content was also found. I think that would be facial recognition for who's on screen. So it's like, oh, you're watching Mr. Beast. That's a fact. No, no, no. I think it's it's watching to see who's watching. I think that might be what's going on. But yeah, the TV, the TV industry has just sort of been able to like, skate by on different rules. Like Apple's done all of this work around privacy and, And even Meta has like, well, yeah, we have camera glasses, but there's going to be a light on. And the TV industry just isn't under the same amount of pressure or scrutiny. So I think they can just... kind of do whatever they want. - Today, Instagram announced they are testing long form horizontal videos called features. The platform wants to be a place where you watch a 90 second video and a 90-minute video all on your TV. They are also testing lives in the Reels tab, a hub for series, and more. So coming straight for... YouTube. I like it. I think more competition is good. I'm interested to see how YouTube responds to this. You know, when Netflix started courting YouTubers, YouTube said, hey, if you put your videos there, we're going to destroy your reach. We'll see what they do here. The feature that I think is most interesting. So, yeah. And Masseri was talking about this on Colin and Samir's show. Cool. Right now, we have a big problem. And we may have played a part in it, but the big problem is that you get a clip, right? And you get this really interesting moment. And we're all looking for the guy who did this. Yeah, we're all looking for the guy who did this. But you get this bit, you get you get this interesting, you know, 30 seconds, let's say. And. Some percentage of the people watch the clip and then go find the full length thing. But most of the people get like they feel some emotion, whether it's joy or rage. Some percentage, like two basis points. Yeah. Yeah. We've we've noticed this where like our our core show, the recording of the live and the diet, they just sort of grow at a steady rate every single day. But like a viral clip here and there doesn't actually drive like a huge uptick. Yes. And so that's just proof that it can drive like a conversation. Wait, that's all they talk about all day? You know, like people can see it as like a-- And so what I think is interesting here And I'm sure that YouTube will respond in some way. It feels relatively straightforward to implement. It's like you'll be able to watch a clip, and then just click into the, like, while in Instagram, click into the main, the full video to that exact point, and then you can, like, watch the whole conversation around it. I think it's going to be very healthy because... Yeah, this is. Yeah, because we've had instances in the past where like someone watches one of our clips, gets angry at it. And like because of something that we said, you know, immediately outside of that, like short little bit. And now we understand that that's like a reality of this situation and that could potentially be bad or good or whatever it is. Yeah. But I think just like more and more competition is good. the question but we didn't clip the question so it looked like the answer was not aware of that or something. I don't know. It happens all the time. Yeah. But anyways, I will be very curious if people start watching Instagram on TV. It's been a great-- Day. It's that time. Yeah, we're so close. We're so close to Chris. We're just hours away. Next week's going to be rough. This week's going to be rough. Leave us five stars. We love you. Bye.
