TWiST News: M.G. Siegler Explains the Current Venture Crisis and Google’s Antitrust Woes | E2023

Episode Summary

Executive Summary: The episode centered on MG Siegler’s essay about the venture capital cycle: too much capital at early stage, broken late-stage liquidity, and how antitrust/regulatory pressure plus weak M&A are freezing the startup ecosystem. The discussion then pivoted to Google’s possible breakup remedies, AI product competition, and how AI could reshape consulting, law, and broader economic distribution.

Main Topics: Venture capital is stuck in a liquidity crunch (Priority: 5/5): MG argues the industry is in a crisis: excess early-stage capital, inflated valuations, and too few exits or acquisitions to recycle capital back to LPs. Early-stage vs late-stage venture dynamics (Priority: 5/5): The hosts and MG contrast early-stage investing’s need for huge multiple outcomes with late-stage funds’ dependence on public-market exits and M&A, both of which are impaired. M&A and antitrust as the system’s release valve (Priority: 5/5): They argue acquisitions—especially tuck-ins and talent buys—are critical for capital recycling and founder/company progression, but regulatory scrutiny has chilled dealmaking. AI boom: excitement now, uncertain winners later (Priority: 4/5): The conversation examines why AI startups command high valuations despite unclear product-market fit and whether incumbents like Apple, Google, and Meta can win by embedding AI into existing distribution. Google antitrust remedies and breakup possibilities (Priority: 5/5): The second half focused on DOJ remedy options for Google, including limits on distribution contracts, data-sharing requirements, and potential divestitures like Chrome or YouTube. AI’s impact on labor and professional services (Priority: 4/5): The hosts discussed how tools like ChatGPT/O1 could reduce demand for consultants and other billable-hour businesses, shifting profits upward while shrinking headcount. Broader social adaptation to automation (Priority: 3/5): The episode closed with speculation about UBI-style systems, child investing accounts, and new models for participation in capital markets as AI changes work and earnings.

Key Arguments: VC returns depend on large exits, but too many companies are overvalued and too few can become fund-returning outcomes. The early-stage market is crowded with capital, making it harder to find 10x-plus opportunities at reasonable entry prices. M&A is not just about big-tech convenience; it is a key mechanism for recycling talent, capital, and founder ambition back into the startup ecosystem. Regulatory limits on acquisitions can freeze the ecosystem by preventing small exits that fund the next generation of startups. AI startups may be overvalued relative to their near-term user businesses, but incumbents still have distribution advantages through phones, browsers, and default placements. Google’s antitrust remedy could revolve around search distribution deals, data access, and possibly structural divestitures rather than an immediate breakup. AI is already replacing some research and consulting work, but consultancies persist because they provide organizational cover and accountability shielding. A future in which corporations become more profitable with fewer employees may require new mechanisms for citizen wealth-building and redistribution.

Data Points: CRV capital return decision: Returning capital to LPs - Referenced as an example of late-stage venture stress and lack of good deployment opportunities. Early-stage return target: 10x - MG described this as a reasonable expectation for strong early-stage venture outcomes. Growth-stage return target: 3x - Used as the rough expectation for later-stage venture funds. Fund return math example: 40x needed to return a 40-company fund - Jason noted that if a fund has 40 names, one company must roughly reach 40x to return the fund. OpenAI scale: 300–400 million monthly active users - Mentioned as evidence of unusually fast consumer adoption for an AI product. OpenAI early valuation: $3 billion+ - Referenced as the valuation around a major early institutional fundraising round. Safe Superintelligence valuation: $1 billion from day one - Used to illustrate how top AI talent can command extreme early-stage pricing. OCI compute savings: 50% less - Oracle promotion claim about compute costs on OCI. OCI networking savings: 80% less - Oracle promotion claim about networking costs on OCI. Gusto customer count: 300,000 businesses - Promotional stat cited for Gusto payroll/HR adoption. Google market revenue context: Tens of billions per year - Described for search and search-text advertising markets in the antitrust case. Google public-device advantage: Billions of devices - Used to describe the inherent distribution advantage of Apple and Google ecosystems. AI query usage example: 30–50 interactions per day - Jason said he uses O1 this many times daily as a workflow replacement.

Pivotal Quotes: "Meta kept telling us to our faces that there was no truth to that whatsoever" — Jason Calacanis: Used to illustrate how major tech companies deny secret projects despite abundant evidence to the contrary. "There’s just so much BS, you know, that’s reported on that is just taken as fact" — MG Siegler: MG explaining why journalism and VC require different approaches to truth, access, and inference. "The biggest one, there’s just been no M&A activity aside from these, whatever you want to call it, hacquisitions" — MG Siegler: Discussing why the startup ecosystem is frozen and why capital recycling is broken.

Implications: If M&A and IPO windows stay shut, startup formation, VC fundraising, and founder recycling will remain constrained. AI may boost winners and productivity, but without exits and policy clarity, the ecosystem could stay distorted and increasingly winner-take-most.