Jason predicts a “major M&A moment” in the next six months! | E2213

Episode Summary

Executive Summary: In this episode of This Week in Startups, Jason Calacanis and Alex Wilhelm discuss NVIDIA's earnings, the AI landscape, GPU depreciation, and the rise of prediction markets. They analyze NVIDIA's strategic investments in OpenAI and Anthropic, debate Michael Burry's fraud accusations regarding hyperscaler depreciation schedules, and highlight Google's surprising lead in consumer AI adoption. The hosts also explore the implications of rapid AI model releases, the potential for major M&A activity, and offer career advice for software engineers in the age of AI.

Main Topics: NVIDIA Earnings and AI Bubble Discourse (Priority: 5/5): NVIDIA beat earnings and guidance, but stock volatility persists. Jensen Huang's comments on AI bubble dynamics and the company's strategic investments in OpenAI and Anthropic are discussed, with skepticism about the firmness of these commitments. GPU Depreciation and Michael Burry's Fraud Accusations (Priority: 5/5): Michael Burry claims hyperscalers are understating depreciation by extending GPU useful lives, potentially overstating earnings by $176 billion by 2028. The hosts counter that extending depreciation from 3-6 years is reasonable given continued utility of older chips. Google's Consumer AI Lead Over OpenAI (Priority: 4/5): Google's Gemini app has 650 million monthly users, and AI Overviews reach 2 billion users, surpassing ChatGPT's 800 million weekly users. This suggests Google is winning consumer AI adoption, challenging OpenAI's dominance. Rapid AI Model Releases and Competitive Landscape (Priority: 4/5): A week of major AI model launches (Grok 4.1, Gemini 3, GPT 5.1, Nano Banana Pro) demonstrates no training wall. Google's Gemini 3 Pro image generation outperforms OpenAI's, highlighting the compounding advantage of focused product development. Prediction Markets and Kalshi's Growth (Priority: 3/5): Kalshi is raising $1 billion at an $11 billion valuation, driven by surging volume from $2 billion in August to $7.5 billion in October. Prediction markets are becoming integral for understanding economic and political outcomes. M&A Outlook and Market Predictions (Priority: 3/5): Jason predicts major M&A activity in the next 6-12 months, with companies valued between $25-250 billion being acquired. Potential deals include Amazon buying DoorDash or Uber merging with Waymo. Career Advice for Software Engineers in AI Era (Priority: 3/5): Jason advises software engineers to focus on building agentic workflows that automate organizational chores, as companies seek to become leaner and more efficient through AI.

Key Arguments: NVIDIA's investments in OpenAI and Anthropic are not guaranteed; they are based on letters of intent and subject to closing conditions, potentially indicating round-tripping concerns. Michael Burry's fraud accusation is overblown; extending GPU depreciation from 3-6 years is reasonable because older chips remain useful for less intensive tasks, and operational costs (power, cooling) provide a natural incentive to retire unprofitable hardware. Google is winning consumer AI adoption due to its integrated ecosystem, with Gemini app users (650M monthly) and AI Overviews (2B users) surpassing ChatGPT's 800M weekly users, though some usage may be passive. The rapid pace of AI model releases (5 major models in one week) shows no training wall, and the compounding effect of AI improving AI tools will accelerate progress further. Prediction markets like Kalshi and Polymarket are becoming essential for understanding real-world probabilities, as evidenced by their rapid volume growth and influence on market expectations. Major M&A is imminent due to favorable regulatory environment and pent-up demand, with potential deals involving mid-cap tech companies in the $25-250 billion range.

Data Points: NVIDIA's potential investment in Anthropic: $10 billion - Subject to closing conditions, announced in November 2025. Gemini app monthly active users: 650 million - Surpassing ChatGPT's 800 million weekly users, indicating Google's lead in consumer AI adoption. AI Overviews monthly users: 2 billion - Google's AI search feature reaches 2 billion users monthly. Kalshi and Polymarket combined volume in October: $7.5 billion - Up from $2 billion in August, driving Kalshi's valuation to $11 billion. Michael Burry's estimated understated depreciation: $176 billion - Between 2026 and 2028, if hyperscalers extend GPU useful lives. Oracle's depreciation schedule change: From 5 years to 6 years - Oracle extended GPU useful life by 20% in 2024. Microsoft's depreciation schedule change: From 3 years to 6 years - Microsoft doubled the useful life of its GPUs.

Pivotal Quotes: "If we delivered a bad quarter, it's evidence that there is an AI bubble. If we delivered a great quarter, we are fueling the AI bubble." — Jensen Huang (paraphrased by Jason Calacanis): Commenting on NVIDIA's earnings and the AI bubble discourse. "Understated depreciation by extending useful life of assets artificially boosts earnings. One of the more common frauds of the modern era." — Michael Burry: Accusing hyperscalers of fraud by extending GPU depreciation schedules. "If that's happening every two to three years, that means theoretically that you might buy more of them. But that does not mean that you're throwing the old ones away. That's ludicrous." — Jason Calacanis: Countering Michael Burry's argument by comparing GPU upgrades to car generations.

Implications: The AI landscape is rapidly evolving with no clear leader, as Google's consumer adoption challenges OpenAI. GPU depreciation debates highlight potential accounting risks in hyperscaler earnings. Prediction markets are becoming essential tools for understanding economic and political outcomes. Major M&A activity is likely in the next year, reshaping the tech industry. Software engineers should pivot to building agentic workflows to remain relevant.