$NVDA's earnings, media economics in the AI era, Farcaster's mega crypto round, and more! | E1954

Episode Summary

Executive Summary: The episode centers on two breaking stories: NVIDIA’s blockbuster earnings, which reinforce its dominance in the AI infrastructure boom, and a $250M OpenAI-News Corp licensing deal that signals a new monetization model for publishers. The hosts debate sustainability, overcapacity, and whether licensing plus AI partnerships can reshape media and startup economics.

Main Topics: NVIDIA’s earnings and AI-driven growth (Priority: 5/5): The hosts unpack NVIDIA’s huge revenue and profit beat, emphasizing how AI demand—especially data center spending for large language models—continues to drive extraordinary growth and shareholder value. Sustainability of AI infrastructure demand (Priority: 5/5): They debate whether NVIDIA’s growth can continue as cloud giants build custom chips and whether the market may eventually hit overcapacity or a plateau in AI compute demand. OpenAI-News Corp licensing deal (Priority: 5/5): The conversation highlights the significance of a quarter-billion-dollar deal for training and content access, framing it as a possible template for publisher compensation in the AI era. Apple News Plus and publisher monetization (Priority: 4/5): The hosts discuss how Apple News Plus is generating meaningful revenue for publishers like the Daily Beast and how curated licensing and subscription distribution may support journalism. Humane AI hardware skepticism (Priority: 4/5): They examine Humane’s reported sale prospects after negative product reviews, using it as a case study in why consumer hardware plus AI is difficult to execute and commercialize. Microsoft Recall, privacy, and surveillance fears (Priority: 4/5): The discussion turns to Microsoft’s Recall feature and startup alternatives like Limitless AI, with strong concern that pervasive recording and cloud storage create unacceptable privacy risks. Farcaster valuation and decentralized social (Priority: 3/5): The hosts question the economics of Farcaster’s large valuation relative to its user base, arguing that decentralized social protocols face weak monetization and network-effect challenges versus centralized incumbents.

Key Arguments: NVIDIA’s scale is extraordinary because it added roughly $19B in revenue year over year, something the hosts compare to historic platform shifts like early Google ads. The AI boom is being fueled by fear and greed among incumbents and cloud providers: they fear losing relevance but also want to capture the upside of AI. NVIDIA’s future upside may be capped by two risks: custom in-house chips from Amazon/Microsoft/Google and eventual overbuilding of AI compute capacity. Publisher licensing deals with OpenAI and Apple News Plus suggest that high-quality content may become a durable paid input for LLMs and AI products. Media monetization is shifting away from ads for many publishers because ad dollars concentrate in Google/Meta, ecommerce marketplaces, and podcasts, making subscriptions/licensing more attractive. Humane’s failure illustrates that consumer AI hardware must solve a real problem better than a phone, at a tolerable price, or it will not survive. Microsoft’s Recall-style tools and always-on recording concepts are seen as powerful but deeply invasive, likely limiting consumer adoption and raising security concerns. Farcaster’s funding/valuation appears disconnected from its current usage because decentralized social is still niche, complex, and unlikely to rival scaled platforms like Reddit or X in the near term.

Data Points: NVIDIA quarterly revenue: $26 billion - Reported earnings for the quarter NVIDIA expected revenue: $24.7 billion - Street expectation referenced during discussion NVIDIA EPS: $6.12 per share - Quarterly earnings per share NVIDIA revenue a year ago: $7.19 billion - Year-over-year comparison cited by hosts NVIDIA after-hours stock move: Up 4.5% - Post-earnings market reaction NVIDIA market cap before earnings: $2.33 trillion - Value going into earnings day NVIDIA data center revenue growth: Up 23% sequentially and 427% year over year - Core growth engine discussed NVIDIA gross margin: 78.4% - Improved quarter over quarter from 76% NVIDIA operating expenses growth: 39% year over year - Used to emphasize operating leverage NVIDIA operating income growth: Nearly 700% year over year - Illustrates profitability expansion NVIDIA next-quarter revenue guide: $28 billion - Management guidance referenced OpenAI-News Corp deal size: $250 million over five years - Licensing agreement discussed Implied annual News Corp deal value: $50 million per year - Host’s breakdown of the deal Apple News Plus revenue for Daily Beast: $3-4 million annually - Semaphore-reported figure discussed NYT Q1 2024 subscribers: 10.6 million total, 9.9 million digital only - Used to benchmark publishing scale NYT Q1 2024 revenue: $557 million - Converted by hosts to roughly $2.2B annualized revenue Farcaster raise: $150 million - Discussed alongside its valuation and user count Farcaster users: 80,000 daily users - TechCrunch figure cited in conversation Farcaster funding round valuation: $1 billion - Hosts question the math of the deal Reddit weekly active users: 306 million - Referenced as a comparison point for social valuation Reddit daily active users: 82.7 million - Used to compare with Farcaster

Pivotal Quotes: "I love breaking news. I love earnings. I love media economics." — Alex Wilhelm: Opening reaction to the episode’s two main stories "We've redistributed wealth from Apple, Google, and Microsoft to NVIDIA." — Jason: Commentary on how AI capex shifts value from platform buyers to chipmaker NVIDIA "Hardware is hard." — Jason: Summarizing why Humane and other consumer AI hardware startups struggle

Implications: AI infrastructure spending is still booming, but its eventual ceiling is uncertain. Meanwhile, content licensing may become the new normal for publishers and AI firms. Startups should expect more AI-driven efficiency, heightened privacy debates, and tougher scrutiny of inflated valuations.