Jason Unpacks Sequoia’s New Funds | E2199

Episode Summary

Executive Summary: The episode pivots from startup chatter to major tech and market signals: hyperscaler earnings as a barometer for AI demand, AI-driven expense fraud, drone and synthetic-data demos, self-driving progress, Sequoia’s new fund structure, Mercor’s revenue quality, AI prediction markets, and dark patterns at Microsoft/Amazon. The hosts repeatedly argue that AI is amplifying both opportunity and misuse, and that founders should pay attention to incentives, transparency, and reputation.

Main Topics: Hyperscaler earnings as an AI demand signal (Priority: 5/5): Meta, Alphabet, Microsoft, and Amazon are framed as the key public-market readout for AI infrastructure spending, compute constraints, and downstream demand for GPUs, cloud, and AI products. AI-enabled fraud and expense management (Priority: 4/5): The hosts discuss stories of employees using AI to forge receipts and how companies are responding with AI detection, while emphasizing that cheating is old behavior in a new form and trust is costly to lose. Drones, live entertainment, and synthetic spectacle (Priority: 3/5): Drone light/fireworks shows in China, a Bitcoin-themed Swiss drone display, and a Vatican drone event are used to explore how drone shows may evolve into a new category of live entertainment and PSYOPS-style communication. Synthetic driving data and autonomy (Priority: 5/5): Tesla’s fully synthetic FSD demo and similar simulator efforts from Wave and Wabi are discussed as ways to accelerate AV training, alongside Uber/NVIDIA data-sharing and Waymo safety stats. Venture fund strategy and Sequoia’s new vehicles (Priority: 4/5): Sequoia’s new seed and Series A funds are analyzed in the context of its permanent-capital model, partner discipline, and what it signals about stage sizing and portfolio focus. Mercor, marketplaces, and revenue definitions (Priority: 4/5): The hosts debate whether Mercor’s revenue should be considered recurring and how to distinguish gross vs. net revenue in marketplace-like businesses that pay out most of the top line. Dark patterns, reputation, and founder ethics (Priority: 4/5): Microsoft’s consumer AI upsell in Australia and Amazon’s Prime cancellation/friction issues are cited as examples of dark patterns, with advice for startups to avoid short-term growth tricks and protect trust.

Key Arguments: AI will increase search and productivity usage rather than replace it, because better answers and automation create more follow-on activity (Jevons paradox). Hyperscaler earnings are the most important near-term public signal for AI demand because their capex and compute constraints reveal real customer appetite. Receipt fraud predates AI; AI simply lowers the barrier, but companies will counter with AI-based detection and trust damage is the bigger issue. Drone shows are likely to become a major live-entertainment format, especially when combined with music, branding, and touring events. Synthetic data generation is becoming central to autonomous driving development, helping models train on rare or dangerous scenarios. Self-driving will likely commoditize across several stacks and players, with the big differentiators being data, hardware, software, and regulatory rollout. Sequoia’s fund split reflects discipline and stage specialization rather than a departure from its permanent-capital strategy. Mercor-like businesses may be high-growth but are better understood as marketplace/project revenue than classic SaaS recurring revenue. Founders should avoid political posting unless they are already insulated; public commentary can harm fundraising, employees, and partnerships. Companies should align incentives around renewals and customer satisfaction, not just initial conversion, to avoid dark-pattern behavior.

Data Points: Hyperscaler cohort: 4 companies - Meta, Alphabet, Microsoft, and Amazon are the key earnings reports the hosts want founders to watch. Top cloud providers: 3 of 4 mentioned - Alphabet, Microsoft, and Amazon are described as the top three cloud providers globally. Amazon workforce: 1.4 million employees - Used to highlight the scale of Amazon and the likely scrutiny over automation/robotics. Amazon share of U.S. jobs: 1% of the country - The hosts say roughly 1% of the country works at Amazon. Sequoia Series A fund: $750 million - New Series A-focused fund raised by Sequoia. Sequoia seed fund: $200 million - New seed/pre-seed fund raised by Sequoia. Mercor valuation: $10 billion - Reported valuation for Mercor in the fundraising discussion. Mercor raise: $350 million - Reported round size in the transcript. Mercor revenue payout ratio: 60% to 70% of top-line revenue - Used to argue that gross revenue may overstate economics for marketplace-style businesses. Waymo miles in Los Angeles: 16 million - Safety and scale data cited by the hosts. Waymo miles in San Francisco: 30 million - Safety and scale data cited by the hosts. Waymo miles in Phoenix: 46 million - Safety and scale data cited by the hosts. Waymo safety improvement: 91% fewer serious injuries - Compared with human-driven baseline, per the discussed safety report. Waymo safety improvement: 79% fewer airbags deployed - Safety report statistic discussed on the show. Waymo safety improvement: 80% fewer injury-causing crashes - Safety report statistic discussed on the show. Waymo safety improvement: 92% fewer pedestrian crashes with injury - Safety report statistic discussed on the show. Waymo safety improvement: 78% fewer cyclist crashes - Safety report statistic discussed on the show. Waymo safety improvement: 89% fewer motorcycle crashes - Safety report statistic discussed on the show. AWS Activate credit offer: Up to $100,000 - Sponsor mention tied to startup infrastructure support. Monarch discount: 50% off first year - Sponsor mention for personal finance management. Pipedrive trial: 30-day free trial - Sponsor mention for CRM. Microsoft consumer AI upsell: Higher-paying tiers in Australia - The hosts discuss a dark-pattern complaint where customers were pushed toward AI-enabled plans. FTC Amazon settlement: $2.5 billion - The hosts cite a historic settlement over Prime enrollment/cancellation issues. FTC Amazon penalty: $1 billion civil penalty - Part of the settlement discussed. FTC Amazon refunds: $1.5 billion - Part of the settlement discussed.

Pivotal Quotes: "They’re your robots taking your job. Period, full stop. Find another job quickly." — Speaker 2: A blunt reaction to Amazon robotics and automation narrative management. "Better answers are better, and if the answers are better and you solve problems quicker, you still have a certain amount of time. You will be more inquisitive." — Speaker 2: Used to argue that AI will expand, not shrink, search and product usage. "If your mom’s not going for it and she’s like, I’m very disappointed in you, then don’t do it." — Speaker 2: Advice to founders about avoiding dark patterns and shady growth tactics.

Implications: Founders should treat AI as both a growth engine and an integrity test: more automation means more opportunity, but also more fraud, scrutiny, and backlash if trust is mishandled. Public earnings, AV safety data, and fund structures now signal where AI is headed.