Jensen Huang Resigns From… | The VC Corner Summary

1. Key Themes

Theme 1: AI mega-valuations are being repriced as private hype meets public markets

Anthropic's IPO range sits well below earlier $2T chatter

Anthropic's proposed pricing implies a valuation far under the figure reported only months ago, which serves as a marker for every late-stage AI mark on a cap table.

"The top of the range sits well below the $2 trillion figure reported a couple of months ago, a useful marker for every late-stage AI valuation still on a cap table."

"Anthropic is offering 825 million Class A shares, which would raise $57.75 billion to $64.35 billion, and is selling $26 billion of non-voting stock at the IPO price to seven tech companies, with Google and Nvidia putting in $10 billion apiece."

OpenAI's round is also coming in smaller than planned

OpenAI is raising less than it originally sought, as investors get pickier on AI pricing.

"OpenAI closed the first $11 billion of a round that could reach around $20 billion, below the $30 billion it first sought."

"...so the gap between private hype and public pricing is narrowing."

Nvidia's leadership shock hit the whole AI complex

The sudden CEO exit erased a huge amount of market cap and raises questions about Nvidia's role as the ecosystem's dominant AI investor.

"Shares were down nearly 12% in early trading on Thursday and had fallen as much as 19% since the prior Friday, more than $855 billion in market cap."

"Nvidia is also one of the most active startup investors in AI, so every portfolio company will watch how the next CEO treats that program."

Theme 2: AI agents are becoming a primary actor on the web, in code, and in commerce

Agents are now meaningful buyers and builders

Agent-driven activity is scaling quickly across e-commerce, software development, and deployment.

"Shopify saw orders from AI-powered searches grow nearly thirteenfold year over year in Q1 2026. Meanwhile, over 40% of GitHub pull requests started with agent-written drafts by September, up from 17%, while Vercel agent deployments increased 28-fold in a year."

Companies aren't prepared for consumer agents

Personal agents are negotiating with customer service, and enterprises can't detect them or route them efficiently.

"Personal AI agents are negotiating bills and chasing refunds, but customer service leaders cannot reliably identify how much inbound traffic they generate. Without dedicated machine-to-machine channels, human support queues absorb the additional volume and concessions at full operating cost."

Enterprise agent ROI is already showing up in banking

Payback periods are short when agentic projects are well-scoped.

"Across 52 banking AI agent use cases, average payback takes around ten months, with a three-year ROI of approximately 2.9x. Leading banks are gaining an advantage through better model context and faster transitions from pilots to production."

Theme 3: The AI buildout is getting physical (power, hardware, real-world data)

Labs are buying power infrastructure rather than renting it

Anthropic's turbine purchase signals where demand is heading for energy and data center hardware.

"Anthropic is buying Caterpillar generators that equipment supplier OnePWR had originally ordered, capable of producing up to 1.7 gigawatts of electricity. A lab that has historically preferred to rent infrastructure is now buying it, which tells founders in energy and data center hardware where the next wave of demand sits."

Value in training data is shifting to novel real-world data

Data is becoming a priced asset, and original, operationally generated data commands a premium.

"Raw data captures only a small share of potential value, while tasks built from it can sell for 2 to 5 times more, favoring businesses generating original data through real-world operations."

Capital is flowing to physical AI and defense

Several of the week's largest rounds target physical-world applications.

"UFORCE raised $700M in Series A funding at a $6B valuation to scale autonomous defence systems."

"Path Robotics raised $300M in Series E funding to scale its physical AI platform for manufacturing."

Theme 4: Venture capital is concentrating into giant rounds while early-stage activity shrinks

First rounds have halved while mega-rounds dominate

"Global first rounds have fallen to 51% of 2022 levels, averaging 5,200 quarterly compared with 10,200. In Q3, 299 rounds of at least $100 million captured 80% of venture funding, pushing the average deal size from $5.7 million to $27.1 million."

Fund formation reflects the same barbell, with large multi-stage funds and niche specialists

"Norwest closed Norwest 18 at $4B, its largest fund yet, a multi-stage fund for technology, healthcare and consumer companies in the US, India and Israel with a deeper push into AI."

"Zacua Ventures closed its second fund at $135M, investing in early-stage technology across construction and physical industries."

Theme 5: AI adoption in security is outrunning AI security readiness

Usage is surging faster than governance

"Active AI use in security strategies rose from 50% to 78%, while 63% of practitioners reported significant shortcomings in AI threat detection and response. Only 36% reported formal AI risk programs, and 78% of organizations had faced confirmed or suspected AI-enabled attacks."

Security startups are drawing large checks

"Daylight raised $80M in Series B funding to scale its AI-powered cybersecurity platform."

2. Contrarian Perspectives

The AI valuation super-cycle may be cooling, not accelerating

While the narrative around frontier labs has been ever-higher marks, the evidence points to price discipline returning. Anthropic's IPO tops out around $1.4T versus roughly $2T earlier expectations, and OpenAI raised less than targeted.

"Anthropic is now targeting a valuation of around $1.4 trillion for its IPO, down from earlier expectations of roughly $2 trillion, so the gap between private hype and public pricing is narrowing."

A frontier AI lab pivoting from renting to owning infrastructure

The conventional view is that labs stay asset-light and rent compute and power. Anthropic's $3.4B turbine purchase cuts against that.

"A lab that has historically preferred to rent infrastructure is now buying it, which tells founders in energy and data center hardware where the next wave of demand sits."

Seed and Series A are shrinking even as AI is "hot"

The boom in AI has not translated into healthy early-stage deal flow. Capital is concentrating at the top, leaving a thinner early-stage market.

"In Q3, 299 rounds of at least $100 million captured 80% of venture funding, pushing the average deal size from $5.7 million to $27.1 million."

3. Companies Identified

Nvidia

Description: AI chip leader and one of the most active AI startup investors.
Why mentioned: CEO Jensen Huang resigned, CFO Colette Kress became interim CEO, and the stock dropped sharply. Nvidia is also investing $10B in Anthropic's IPO.
Quotes: "Nvidia's board accepted Huang's resignation on November 11, named Kress interim CEO while she stays CFO, and made longtime director Harvey C. Jones its first-ever board chair." / "Google and Nvidia putting in $10 billion apiece."

Anthropic

Description: Frontier AI lab preparing a major IPO.
Why mentioned: IPO range of $70 to $78 a share and a $3.4B gas turbine purchase.
Quotes: "Anthropic sets its IPO range at $70 to $78 a share, implying a valuation of up to $1.404 trillion" / "Anthropic is buying Caterpillar generators that equipment supplier OnePWR had originally ordered."

OpenAI

Description: Frontier AI lab raising private capital.
Why mentioned: Closed the first $11B of a round at a $1.25T valuation, smaller than originally sought.
Quotes: "OpenAI closes the first $11bn of its round at a $1.25tn valuation as investors get pickier on AI pricing"

Google

Description: Tech giant investing in Anthropic.
Why mentioned: Committing $10B to Anthropic's IPO.
Quotes: "Google and Nvidia putting in $10 billion apiece."

Caterpillar / OnePWR

Description: Generator manufacturer and the equipment supplier that originally ordered the turbines.
Why mentioned: Source of the 1.7GW of generators Anthropic is buying.
Quotes: "Anthropic is buying Caterpillar generators that equipment supplier OnePWR had originally ordered, capable of producing up to 1.7 gigawatts of electricity."

Horizon3

Description: Autonomous security testing company.
Why mentioned: Case study in rapid scaling, from $1M to $100M ARR in five years, and a channel-led go-to-market.
Quotes: "Horizon3 reached $100M ARR in October 2026 with 95% gross retention and 125% net retention, building 80% of its first 3,000 customers through channel partners."

Shopify

Description: E-commerce platform.
Why mentioned: Evidence of agent-driven commerce growth.
Quotes: "Shopify saw orders from AI-powered searches grow nearly thirteenfold year over year in Q1 2026."

GitHub / Vercel

Description: Developer platforms.
Why mentioned: Indicators of agent-driven software development.
Quotes: "over 40% of GitHub pull requests started with agent-written drafts by September, up from 17%, while Vercel agent deployments increased 28-fold in a year."

Metaview (fillmore)

Description: Recruiting software with a Slack-native agent.
Why mentioned: Example of an outbound recruiting agent.
Quotes: "Teams brief fillmore by Metaview in Slack with a job description or sample profiles, and the agent researches prospective candidates and writes tailored outreach to book screening calls directly on the calendar."

Attio

Description: CRM, sponsor of the issue.
Why mentioned: Promotes an agentic sales audit.
Quotes: "Attio's agentic sales audit looks at where you are today and shows you where to start, whether that's your data foundation or a fuller agentic sales process."

Norwest

Description: Multi-stage venture firm.
Why mentioned: Closed its largest fund, Norwest 18, at $4B.
Quotes: "Norwest closed Norwest 18 at $4B, its largest fund yet."

Mercor

Description: AI training and enterprise infrastructure company.
Why mentioned: Raised $750M Series D at a $20B valuation.
Quotes: "Mercor raised $750M in Series D funding at a $20B valuation to expand its AI training and enterprise infrastructure."

UFORCE

Description: Autonomous defence systems company.
Why mentioned: $700M Series A at a $6B valuation.
Quotes: "UFORCE raised $700M in Series A funding at a $6B valuation to scale autonomous defence systems."

Path Robotics

Description: Physical AI platform for manufacturing.
Why mentioned: $300M Series E.
Quotes: "Path Robotics raised $300M in Series E funding to scale its physical AI platform for manufacturing."

Imprint

Description: Credit card and financial products company.
Why mentioned: $130M growth round led by Khosla Ventures.
Quotes: "Imprint raised $130M in growth funding, led by Khosla Ventures, to expand its financial products and credit-card business."

Databento

Description: Financial market data platform.
Why mentioned: $100M Series C at a $1.1B valuation.
Quotes: "Databento raised $100M in Series C funding at a $1.1B valuation to expand its financial market data platform."

Rox

Description: AI-powered enterprise revenue platform.
Why mentioned: $100M Series B at a $1.3B valuation.
Quotes: "Rox raised $100M in Series B funding at a $1.3B valuation to expand its AI-powered enterprise revenue platform."

Daylight

Description: AI-powered cybersecurity platform.
Why mentioned: $80M Series B.
Quotes: "Daylight raised $80M in Series B funding to scale its AI-powered cybersecurity platform."

Charta Health

Description: AI medical chart review platform.
Why mentioned: $55M Series B.
Quotes: "Charta Health raised $55M in Series B funding to expand its AI-powered medical chart review platform."

Oyster Therapeutics / Eli Radiopharma

Description: Biotech companies in protein degradation and radiopharmaceuticals.
Why mentioned: Early-stage biotech rounds.
Quotes: "Oyster Therapeutics raised $60M in Series A funding to advance its targeted protein degrader pipeline." / "Eli Radiopharma raised nearly RMB 400M in Series A funding to advance its radiopharmaceutical pipeline."

Newly closed funds (redalpine, Vine Ventures, Zacua Ventures, Stage 2 Capital, Course Corrected, Proximal Ventures, Noctua Science Ventures)

Description: Specialist and regional venture funds.
Why mentioned: Fund formation data points showing thesis-driven niche strategies.
Quotes: "redalpine closed its eighth early-stage venture fund at $250M" / "Stage 2 Capital closed $100M across Venture Fund IV and Catalyst IV to back B2B AI companies." / "Proximal Ventures closed its debut fund at $27.5M to invest in AI-driven life sciences companies."

4. People Identified

Jensen Huang

Description: Nvidia's CEO, who resigned.
Why mentioned: The week's headline leadership change.
Quotes: "Nvidia's board accepted Huang's resignation on November 11"

Colette Kress

Description: Nvidia CFO, now interim CEO.
Why mentioned: Taking over operational leadership while remaining CFO.
Quotes: "named Kress interim CEO while she stays CFO"

Harvey C. Jones

Description: Longtime Nvidia director.
Why mentioned: First-ever Nvidia board chair.
Quotes: "made longtime director Harvey C. Jones its first-ever board chair."

Dr. Z. Berkay Celik

Description: Purdue researcher.
Why mentioned: Leads Horizon3's new Adversarial Intelligence Frontier Lab.
Quotes: "Horizon3's Adversarial Intelligence Frontier Lab, led by Purdue's Dr. Z. Berkay Celik, turning attack data from 7,000+ organizations into automated attack-and-defense agents."

Andre Retterath

Description: Author credited with the training-data analysis.
Why mentioned: Source of the insight on training data as a priced asset.
Quotes: "Raw data captures only a small share of potential value, while tasks built from it can sell for 2 to 5 times more." [Andre Retterath]

Ruben Dominguez

Description: Author of The VC Corner.
Why mentioned: Curator of the newsletter and its resources.
Quotes: "Another week, another pulse check on Tech."

5. Operating Insights

Secure agent access to data at the database layer

The security boundary is the database credential, not the agent prompt. Give each agent its own identity with minimal roles, cap outputs, and convert recurring questions into reviewed parameterized queries.

"The guide argues the real security boundary is the database account an MCP server logs in with, so each agent needs its own identity and only the roles its tasks require. It caps results at 25 rows by default and 500 per query, clips each cell at 500 characters, and moves recurring business questions to reviewed, parameterized queries so answers stay repeatable."

Use channel partners to scale go-to-market

Horizon3's growth was largely partner-driven, with strong retention backing it up.

"Horizon3 reached $100M ARR in October 2026 with 95% gross retention and 125% net retention, building 80% of its first 3,000 customers through channel partners."

Fix data and process foundations before scaling agents in sales

Agents are only as effective as the data and processes behind them, so audit these before scaling headcount.

"You're growing from 10 reps to 100. Your agents can only work with the data and processes you give them, and the foundations you set now decide whether your systems hold."

6. Overlooked Insights

Customer support teams are absorbing hidden agent traffic at full human cost

Companies may be subsidizing consumer agents' bill negotiations and refund claims without being able to see it, a gap that suggests demand for agent-identification and machine-to-machine support tooling.

"Without dedicated machine-to-machine channels, human support queues absorb the additional volume and concessions at full operating cost."

Strategic investors are anchoring Anthropic's IPO

A large slice of the offering is pre-placed with strategic buyers, signaling that non-voting, cornerstone-style demand from hyperscalers and chipmakers underpins the deal.

"...is selling $26 billion of non-voting stock at the IPO price to seven tech companies, with Google and Nvidia putting in $10 billion apiece."