VCs Debate AI Bubble While Literally Creating It

In what may be the most transparent admission of cognitive dissonance in venture capital history, Connie Loizos convened a live panel at StrictlyVC LA featuring Basis Set Ventures partner Chang Xu and M13 founder Carter Reum to discuss whether today's AI valuations represent an actual bubble. The very existence of this conversation—recorded for public consumption, no less—confirms what observers have known for eighteen months: the people writing the checks are genuinely unsure if they're participating in a market correction or a mass delusion event, and they're uncomfortable enough to monetize the uncertainty through a podcast appearance.

The structural irony here deserves unpacking. These panelists manage capital that has directly funded AI companies now trading at valuations disconnected from revenue, customer acquisition costs, or path to profitability. Xu and Reum aren't external commentators observing a bubble from safe distance—they're seated inside the bubble, funded by LPs who expect returns from the very companies whose valuations they're now publicly questioning. This is not due diligence. This is performance art masquerading as market analysis, staged in El Segundo for an audience of limited partners and aspiring founders who need permission to keep writing bigger checks.

The topic itself—"ARR inflation"—deserves scrutiny. When a VC partner discusses how startup revenue figures may be misleading or overstated, he is discussing a problem his own firm likely overlooked during term sheet negotiation. No venture firm conducts reverse diligence to identify where they might have been wrong about a valuation. Instead, they host panels to discuss the methodology that allowed them to be wrong in the first place, framing systemic failure as intellectual curiosity. The podcast format ensures they can explore these doubts without committing to any actual positions or accountability.

What makes this particularly rich: the conversation exists because the question is no longer rhetorical. If AI valuations were obviously justified by fundamentals, there would be nothing to debate. The fact that three respected capital allocators felt compelled to book studio time in Los Angeles to discuss "whether there's a bubble" is itself confirmation that the bubble exists. Markets operate on confidence. Once that confidence requires public panel discussion to reinforce it, it's cracking.

Consider what could happen next: a correction. When it arrives—and corrections are statistically inevitable—these same VCs will claim the problem was irrational exuberance among "other investors," not structural issues with their own thesis or process. They'll cite the StrictlyVC panel as evidence they were thinking critically all along, despite the fact that thinking critically would have meant deploying capital more conservatively, not hosting a podcast about your doubts while your checks continue to clear.

The broader pattern here is familiar. Every cycle produces its self-aware bubble panel: late 2000s had housing experts discussing housing bubbles; 2018 had blockchain investors discussing blockchain bubbles. The presence of doubt at the top of a market is not a sign of healthy skepticism—it's a sign that returns are already being priced for disappointment, and the people responsible are hedging their reputational exposure through media appearances.

If Xu and Reum genuinely believed AI valuations were inflated, they would deploy capital differently. Instead, they booked a panel. Some conversations are just expensive ways of admitting you're complicit.