The Twenty Minute VCAnthropic’s $10B Raise | a16z’s $15B Fund: Is the Middle Dead in VC? | How OpenAI Could Go to Zero?
At a glance
WHAT IT’S REALLY ABOUT
AI mega-rounds reshape venture capital: scale, risk, and policy shocks
- Anthropic’s reported $10B raise is framed as potentially pre-IPO and arguably rational on forward revenue math if current extreme growth persists another year.
- Claude’s momentum is discussed as a platform expansion from enterprise API dominance into coding (Claude Code) and potentially broader “knowledge work” via workspace-style products, raising bundling and dependency risks for partners like Cursor.
- OpenAI is argued to have meaningful existential risk only under a compounded scenario where capital access tightens, model half-lives stay short, and consumer/enterprise usage proves more fickle than assumed.
- Andreessen Horowitz’s $15B fundraise is presented as a system-level scaling strategy: win ~10% of the best early deals, then concentrate heavily in winners later to offset higher miss rates at scale.
- The group debates whether VC is bifurcating into mega-platforms and focused boutiques while the “middle” gets squeezed, and concludes policy (wealth tax) plus AI-driven labor/wealth divergence could materially alter where companies are built and who stays in California.
IDEAS WORTH REMEMBERING
5 ideasAnthropic’s price can look cheap if one more growth year holds.
The hosts model Anthropic’s run-rate trajectory (100M → 1B → ~10B → potential ~30B) to argue a ~$350B valuation could be ~17x next-twelve-month revenue—defensible only if growth remains exceptional near-term.
Claude’s strategy is to climb the stack from API to “work operating system.”
They describe three layers—enterprise API, coder product (Claude Code), and a workspace for non-coders—aiming to capture more of the value chain and compete with Office-like bundles rather than being “just an API.”
Partner ecosystems like Cursor face “supplier can become competitor” risk.
Because Cursor depends on frontier models, Anthropic could limit access, degrade top tiers, or replicate adjacent products; the discussion frames this as the scorpion-and-frog dynamic inherent to platform dependency.
OpenAI doesn’t go to literal zero, but can be structurally vulnerable.
One bear case hinges on short model half-life plus massive future capital needs; if a macro shock hits during a fundraising window, competitors with stronger cashflows/margins could keep advancing while OpenAI stalls and usage migrates.
Mega-funds work by engineering a portfolio system, not by perfect picking.
Rory argues that at scale you can be “promiscuous at the A” because later-stage check-writing and concentration in breakout winners can cover early mistakes—shifting the core risk from business selection to valuation exposure.
WORDS WORTH SAVING
5 quotesIn the early stage, you're taking uncorrelated business risk, and in the late stage, you're taking 100% correlated valuation risk.
— Rory O’Driscoll
I think OpenAI has existential risk. It is a bet that, that, that the best of times lasts at least a decade.
— Jason Lemkin
Would you use ChatGPT from a year ago? Would you use Claude from a year ago? NFW you would use these products from a year ago.
— Jason Lemkin
You can be promiscuous at the A if you have enough late stage stuff to cover it up.
— Rory O’Driscoll
This is a Trojan horse. The, this is not about a one-time 5% wealth tax. The goal of the proponents of this bill... is that this will then transition to an annual tax.
— Jason Lemkin
High quality AI-generated summary created from speaker-labeled transcript.
