Podcast Summary – This Week in Startups

Episode: Jason Unpacks Sequoia’s New Funds (E2199)
Host: Jason Calacanis
Date: October 28, 2025

Episode Overview

This episode of This Week in Startups sees Jason Calacanis and co-host unpack major trends in tech, AI, and venture capital, centering on Sequoia Capital’s new funds. The duo also riff on AI’s disruption of everything from corporate compliance to self-driving cars, cultural differences in drone shows, business ethics (with jabs at Microsoft and Amazon), and the primal importance of reputation in startups. As always, there are lively debates, candid opinions, and behind-the-scenes stories from Silicon Valley.

Key Discussion Points & Insights

Why Public Earnings Matter for Startups

• [00:00-02:20]
  • Jason cautions startup founders to pay attention to upcoming earnings from hyperscalers: Meta, Alphabet (Google), Microsoft, Amazon (“MAMA”).
  • These corporations are pouring billions into AI infrastructure; their commentary signals the direction and health of the entire AI market.
  • “If you pay attention to no public markets whatsoever, and as a founder, I get it … but you may want to tune into a couple of these because there’s going to be interesting notes about industries that are buying.” — Jason (14:04)

Insight:

• The feedback loop between big tech’s AI spending and the startup ecosystem is immediate, especially for infra and SaaS founders. NVIDIA and OpenAI continue to benefit as AI demand remains insatiable.

The AI Arms Race—Receipts, Fraud, and Expense Reports

• [07:24-13:23]
  • AI tools are being used by employees to create fake expense receipts; companies like Ramp are countering with their own AI fraud detection.
  • This triggers an “AI arms race inside of companies between cheats and management,” with fraud evolving just as fast as the tools to catch it.
  • “This AI receipt fraud is really just a remix on a very, very old [game]. … Old as time.” — Jason (13:23)

Jason’s Take:

• Policy suggestion: Consider stipends or prepaid corporate cards to limit friction and temptation. “Don’t cheat. It makes you look stupid.”

AI Hype in the Cloud—Earnings Signals for the Future

• [14:04-18:37]
  • The importance of MAMA’s earnings is reiterated, especially regarding AI infrastructure investment, hiring freezes, and automation (e.g., Amazon robotics displacing 1% of US jobs).
  • On Google: Despite predictions of search’s demise, usage and ad revenue rise as AI enables users to get more—and better—answers, creating even more demand (Jevons Paradox at play).
  • “If the answers are better and you solve problems quicker, you still have a certain amount of time. You will be more inquisitive.” — Jason (15:13)

Action Item:

• The show will use AI tools to analyze how often “AI” pops up in earnings calls, reflecting the saturation and hype in Wall Street and public tech narratives.

Live Drone Shows: East vs. West

• [19:52-29:34]
  • Viral video of a Chinese drone show disaster sparks discussion about the sheer scale and ambition of Chinese light displays versus more irreverent western shows (like a Bitcoin meme drone show at a Switzerland conference).
  • Cultural Difference: Chinese shows display state pride and technological might (setting Guinness records). Western versions, though smaller, “have more freedom of speech … and I’ll take that.” — Co-host (26:23)
  • Jason predicts drone shows will revolutionize live entertainment, potentially replacing fireworks: “I would not be surprised if … you could take an existing film like Star Wars, put it through an 8-bit or a 16-bit converter, and then watch Star Wars, the entire full film in the sky.”

Autonomous Vehicles—Synthetic Data & The Race for Safety

• [29:34-41:07]
  • Tesla now trains its FSD (full self-driving) software on highly realistic, AI-generated (“synthetic”) video, not real-world footage. This vastly accelerates learning for rare scenarios.
  • “They now have the ability to build situations for the car to navigate and then test it over and over… it’s extraordinary.” — Jason (29:47)
  • Other AV companies (Wave, Wabi, Waymo) are leveraging synthetic data and massive real world datasets (Uber partners with NVIDIA) to close the safety gap.
  • Waymo’s public safety data: huge reduction in accidents vs. human-driven cars in LA and SF, but the media seems more focused on Tesla mishaps ("Elon for clicks").

Insight:

• FSD progress will likely commoditize—expect pervasive self-driving software as “a deployment issue and a regulatory issue.”

Startup Funding Dynamics – Sequoia’s New Funds and the Permanent Capital Model

• [44:34-55:00]
  • Sequoia raises a $750M Series A fund and a $200M seed/pre-seed fund. This, despite their recent “permanent capital” pivot as an RIA.
  • Jason explains: They maintain both models to satisfy differing LP appetites; funds remain relatively modest compared to the mega-funds of rivals.
  • Sequoia’s discipline is in check size, fund size, and partner capacity (each investor does only three deals/year).
  • “Lower fund size, higher expectations, returns, more concentrated positions.” — Alex (49:25)

Notable Theme Focus:

• AI apps (voice, video), cybersecurity, observability, commerce plus AI (potential $1T+ vertical), and regionally, a renewed interest in Israeli and European founders.

Controversy and Caution—Politics, Social Media, and VC Risks

• [51:30-55:58]
  • Discussed: Sequoia partner Sean Maguire’s controversial social media comments, resulting in high-profile internal fallout.
  • Jason’s advice: VCs and founders are increasingly polarizing, and “founder friendliness” across political lines is no longer a given.
  • “You may not be in that position, and it could cost you your funding, it could cost you your employees, it could cost you partners. Why would you even do that?” — Jason (52:58)
  • Paul Graham (YCombinator) and others can take public stands due to wealth/influence; not true for ordinary founders/executives.

ARR & Marketplace Math – Mercor, Micro1, and Gross vs. Net Revenue

• [57:49-66:08]
  • Mercor (RLHF labor for AI firms) rockets to $100–500M ARR, raising at a $10B valuation. But, they pay out 60-70% to labor—raising the classic gross/net revenue debate.
  • Marketplace model is reminiscent of Uber, YouTube, Airbnb—where most revenue is passed through—but calling this “recurring” can be misleading.
  • “If you wanted to call it recurring revenue … yeah, that might be dishonest or not accurate…” — Jason (61:14)
  • Market for data labeling and RLHF will persist: as AI gets smarter, human expert feedback only gets more important.

AI Trading Bots & Polymarket

• [66:14-68:47]
  • AIs from different countries now compete trading crypto. Chinese LLMs (Deep Seek, Alibaba) outpace their American rivals—maybe “less constrained,” giving an edge?
  • New layer of speculation: people are betting on which AI will win—a meta-wager on the future of machine intelligence.

Ethics, Dark Patterns, and Business Reputation

• [69:07-80:13]
  • Microsoft and Amazon called out for “dark” retention patterns (tricking users into premium upgrades, making cancellations difficult).
  • FTC’s $2.5B fine against Amazon over Prime renewals is cited as a warning.
  • Jason’s philosophy: “You have to be obsessed with your reputation at all times in your life. … if your mom saw it, what would your mom say?” — Jason (77:11)
  • “Be the white wizard. … Take the opposite approach. Be super generous.” — Jason (75:37)
  • For startups: Build alignment in incentives and optimize for long-term retention, not just short-term tricks.

Notable Quotes & Memorable Moments

• On the AI hype cycle:
  • “We’re now in an AI arms race inside of companies between cheats and management.” — Jason (07:24)
• Amazon and automation:
  • “They're not cobots. They're your robots. They're robots taking your job, period.” — Jason (15:13)
• On business incentives:
  • “Show me an incentive. I show you the outcome.” — Jason (77:11)
• Startup reputation:
  • “Don’t cheat. It makes you look stupid.” — Jason (13:23)
• Political commentary for founders:
  • “If you think chiming in on one of the oldest and most bitter conflicts in the world … you’re going to make some incredible progress with your opinion? You’re not.” — Jason (55:35)

Segment Timestamps

• [00:00-02:20] – Public earnings & AI infrastructure: why founders should care
• [07:24-13:23] – AI-enabled expense fraud and the eternal workplace grift
• [14:04-18:37] – Big Tech earnings & the cloud providers’ duality with AI
• [19:52-29:34] – Drone shows: Innovation, culture, entertainment, and the future
• [29:34-41:07] – Self-driving cars, synthetic data, and the commodification horizon
• [44:34-55:00] – Sequoia’s new funds, RIA structure, and fund dynamics
• [57:49-66:08] – Mercor vs. Micro1, ARR debates, and growth in the AI labor market
• [66:14-68:47] – AI bots trading crypto, speculation upon speculation
• [69:07-80:13] – Microsoft/Amazon’s “dark patterns”, and how reputation is built (or lost)

Final Thoughts & Takeaways

• Sequoia’s new funds are smaller and more focused—a sign of discipline after VC excess in recent years.
• Markets and AI are inseparable; hyperscaler spending is a leading indicator for both startups and investors.
• Business ethics and reputation matter more than ever in the era of social media, regulatory oversight, and AI skepticism.
• Cultural and technological innovation (from AI to drone shows) is increasingly shaped by both global competition and community values.

For founders: Stay obsessed with reputation, invest in customer trust, and use ethical incentives. The AI tidal wave is here—surf it wisely and watch the signals coming from both the giants and the upstarts.