The Twenty Minute VCPeter Thiel and Softbank Sell NVIDIA - Why? & Why VC Will Hit $1TRN and The Opening of Retail
CHAPTERS
Cursor’s $2.3B raise at $29B: why this valuation can make sense
The group dissects Cursor’s massive financing and what it signals about agentic coding as one of the strongest AI use cases. They argue the upside comes from rapid adoption, strong product performance, and potentially defensible distribution—while acknowledging the emotional whiplash the deal creates for early-stage investors.
TAM explosion: from “developer tool” to trillion-dollar spend category
They debate the real market size by estimating global developer counts and likely annual spend per developer. The conversation broadens to include non-developers building products via tools like Replit and Lovable, implying a second TAM beyond professional engineers.
The two real risks: profitability (token costs) and durability (competition/platform risk)
After agreeing revenue and growth look compelling, they focus on the two core bear cases: whether the business can be highly profitable and whether it can keep its position. A key tension is that model providers can also be competitors, creating platform/supplier risk.
Switching costs vs rapid model progress: will users churn when the next model ships?
They debate whether ongoing model leaps will keep users switching, or whether coding performance is starting to “asymptote,” allowing tools to retain users through memory and workflow customization. The underlying question: when does the market “congeal” and shares stabilize?
Deflation and price wars: SaaS stickiness vs commodity DRAM-style collapse
They distinguish between benign ‘more value for same price’ deflation and destructive price wars. The group explores a worst-case scenario where AI tooling becomes commoditized and pricing collapses, especially if prompts/knowledge become portable and switching costs fall.
Ramp’s jump from $13B to $32B: late-stage as trading, not company-building
Using Ramp and Cursor as examples, they discuss the accelerating cadence of step-up rounds within the same year. The conversation shifts to whether late-stage investing has become a ‘private public market’ where the edge is trading liquidity events on the way up—while warning liquidity disappears on the way down.
Thiel/SoftBank selling NVIDIA: not the signal—watch the credit markets instead
They argue Thiel’s sale is too small to be meaningful, and SoftBank’s move is more about recycling into higher-risk bets like OpenAI. Tomasz flags more concerning indicators in credit and consumer stress that could foreshadow a sharp repricing across AI infrastructure trades.
AI infrastructure at redline: concentration risk, CapEx surge, and “fast & brutal” corrections
They outline how hyperscaler demand looks strong, but the system is running at extreme speed with high leverage and concentration risk. The main fear isn’t slow decay—it’s a sudden wobble that triggers a rapid, violent correction, especially if inference utilization disappoints.
Early-stage exuberance vs public-market fear: YC heat and founder leverage
Harry describes a frothy early-stage environment with rapid fundraises and founders dictating terms, seemingly disconnected from public market anxiety. Rory argues cycles flip bargaining power, and the best response is to stay human while accepting market realities.
VC to $500B–$1T: concentration, definitions, and what really drives asset growth
They debate whether venture capital will balloon by 2030, noting today’s capital is heavily concentrated in a few mega-winners. The conclusion: asset-class growth depends less on ‘venture’ broadly and more on whether a handful of dominant companies deliver massive outcomes.
Why the best companies won’t IPO: access premium, secondary liquidity, and cheaper private capital
Using Stripe’s tender and the high cost of IPOs, they argue private markets now offer cheaper capital and sufficient liquidity for elite companies. The traditional ‘illiquidity discount’ may have flipped into an ‘access premium,’ reducing the incentive to list publicly.
Opening of retail into venture: tsunami potential—and the painful meeting after losses
They explore regulatory and product changes that could route 401(k)/retail flows into venture via funds, ETFs, and secondaries. Rory warns that long feedback loops and illiquidity can trap retail investors in bad vintages, creating future reputational and ethical pressure.
Quick-fire predictions: valuations, entry price logic, and OpenAI IPO timing
In rapid Q&A, they choose between hot assets and explain when entry price matters versus when ‘winning’ matters. They also give a rough timeline for an OpenAI IPO and briefly discuss incentive distortions when leaders aren’t dilution-sensitive.
