The Twenty Minute VCAnthropic Raises $30BN at $900BN Price | SpaceX Files S1: How Does it Trade | Cerebras Smashes Day 1
CHAPTERS
AI hype meets political reality: why America is souring on the AI wave
The episode opens with a blunt framing: outside of tech hubs, the AI boom is not broadly popular. The hosts set up a recurring theme—brilliant technologists often communicate poorly and invite backlash, while savvy political actors can weaponize public resentment.
Anthropic targets a $900B valuation and hires Andrej Karpathy
The panel reacts to two headline items: Anthropic’s prospective mega-raise and Karpathy joining. They discuss why elite talent moves matter for perception and why Anthropic can effectively set terms given demand.
Do ARR multiples still work for frontier AI—or is it all DCF now?
Rory and Jason argue about how investors are underwriting Anthropic at enormous scale. They claim late-stage Anthropic rounds look ‘post-IPO’ with reduced existential risk, making valuation risk the primary variable.
Why raise $30B if the deal is so attractive to investors? Capex, burn, and balance-sheet war
They argue the raise is rational because compute commitments are massive and the capital de-risks another year of burn. Anthropic can give up a small slice of equity to secure runway and negotiating leverage with hyperscalers spending on its behalf.
Founder strategy contrast: Anthropic’s ‘low-drama cash’ rounds vs OpenAI’s structured mega-deals
Jason posits Dario runs fast, clean financings (even at a perceived discount) to avoid drama, while OpenAI’s deals involve complex contingencies. The panel contrasts simplicity (cash in, done) with heavily structured commitments tied to future events.
The real cost of tokens: Salesforce’s $300M spend and what it implies for AI TAM
They break down Benioff’s claim that Salesforce spent $300M on Anthropic tokens, mostly for coding. The math shows it’s ‘normal’ per engineer, but the macro implications are huge: to justify frontier AI valuations, token spend may need to consume a meaningful share of payroll.
Bear case on tokens: agents can be cheaper than expected (Klaviyo example)
Jason shares how his team runs many agents with modest monthly spend, and recounts Klaviyo’s internal agent framework and surprisingly low agent operating costs. The implication: efficiency gains and better tooling could reduce required token volumes, challenging lofty revenue forecasts.
Public SaaS rebounds: Datadog and Figma re-accelerate—but 2021 multiples won’t return
They review strong quarters (Datadog, Figma) and argue the market has shifted permanently from ‘prospects’ valuation to more traditional revenue/growth/cash-flow metrics. SaaS can recover operationally, but attention and peak multiples have moved to AI.
AI threatens web builders: why Wix/Squarespace look ‘terminal’ and Shopify squeezed the middle
Wix’s decline sparks a broader discussion on traditional website builders being attacked from two sides: AI ‘vibe coding’ tools on one end and Shopify’s dominance in commerce on the other. They also critique buybacks when businesses face structural disruption.
Compute starvation and the hardware boom: Nebius/CoreWeave, data centers, and the bubble question
They debate whether skyrocketing growth in compute providers is justified by scarcity or reflects an unsustainable capex frenzy. A key argument: permitting and build constraints might prevent oversupply and ‘save’ the market from overbuilding.
Cerebras IPO surges: what it signals (and doesn’t) for the IPO window
Cerebras’ strong first day is treated as an ‘N of one’—a rare intersection of hot category (semis/inference), marquee demand, and a clear AI exposure proxy. They caution that buying after the pop has poor base-rate returns, and doubt it broadly opens the IPO door for smaller names.
SpaceX S-1 and the $1.75T IPO rumor: excitement, incomplete financial story, and retail dynamics
They anticipate SpaceX’s filing and argue the S-1 may not reflect major recent AI-related acquisitions/deals, forcing storytelling into the roadshow. They debate whether retail enthusiasm can drive a meme-like surge from an already massive valuation, and note price-target dynamics could create volatility.
YC’s OpenAI token-for-equity offer: tokens as marketing, valuation anchoring, and surplus-capacity signal
The panel analyzes OpenAI offering YC startups token credits in exchange for equity. They argue it can inflate valuations via anchoring, shrink round sizes, and encourage aggressive freemium/token-heavy go-to-market—while also implying OpenAI has surplus capacity (vs Anthropic).
Elon vs OpenAI lawsuit aftermath and the coming political backlash from layoffs
They revisit the dismissal of Musk’s OpenAI case (statute of limitations) and argue the broader effect is continuing legal/PR pressure on OpenAI. The episode ends on a warning: mass layoffs plus careless AI messaging will intensify public hostility and drive political responses, potentially forcing tech to ‘re-inflate’ hiring to maintain social stability.