The signal in the noise

Mark Cuban and Michael Burry’s comparison of today’s market to the dot-com bust is not hyperbole. It is a data point that demands attention from venture capitalists, founders, and limited partners alike. Their warning centers on the AI-adjacent sector, where valuations have detached from fundamentals, and liquidity is increasingly scarce. For funds, this is less about predicting a crash and more about preparing for a correction that could reshape the private markets landscape.

The parallels are striking. In the late 1990s, capital flooded into unproven internet companies with little revenue or clear business models. Today, AI startups — many with similar traits — are attracting outsized valuations despite uncertain monetization paths. The question is not whether a correction will come, but how severe it will be and who will be most exposed when it does.

What the warning flags look like in practice

Burry and Cuban’s concerns are not isolated. They point to four specific red flags in the AI-adjacent sector:

Valuation multiples detached from fundamentals

The median revenue multiple for AI startups has climbed to 25x, a level last seen in 2000. For comparison, the S&P 500’s median multiple is 22x, and even high-growth SaaS companies trade at 12x. This disconnect suggests that investors are pricing in future dominance rather than current performance.

A surge in “zombie” AI startups

A growing number of AI companies are raising capital at high valuations but failing to achieve profitability or meaningful revenue growth. These “zombies” are kept alive by investors hoping for a miracle exit, but their existence distorts the market by crowding out healthier competitors.

Liquidity crunch in secondary markets

Secondary market activity for AI startups has plummeted by 40% year-over-year, according to PitchBook. This lack of liquidity means that funds holding these assets may struggle to exit, forcing them to mark down valuations or extend holding periods. For LPs, this translates to longer lock-up periods and potential write-downs.

The rise of “AI-washing”

A proliferation of startups claiming to leverage AI without tangible proof has led to a market where hype often outweighs substance. Investors are increasingly skeptical of startups that cannot demonstrate clear AI differentiation or defensibility. This skepticism could lead to a sharp repricing of assets.

Why this matters for venture funds

For venture funds, the warning signs are a call to action. The most immediate risk is the potential for mark-to-market losses on AI-adjacent portfolios. Funds that entered the space late or overpaid for assets could face significant pressure to write down valuations. This is particularly acute for early-stage funds, which often lack the reserves to weather prolonged downturns.

Portfolio construction in an uncertain environment

Funds should reassess their exposure to AI-adjacent sectors. This does not mean abandoning the space entirely, but it does require a more disciplined approach:

  • Focus on fundamentals. Prioritize startups with clear revenue traction, defensible moats, and realistic unit economics. Avoid companies that rely solely on hype.
  • Diversify across stages. Early-stage funds should consider shifting capital toward later-stage companies with proven business models, while growth-stage funds may need to tighten underwriting standards.
  • Build in downside protection. Structuring deals with ratchets, anti-dilution clauses, or milestone-based funding can mitigate the risk of overpaying for assets.

Fundraising and LP expectations

For funds raising capital, the current environment could make it harder to secure commitments. LPs are increasingly scrutinizing AI exposure in portfolios and may demand more conservative assumptions about returns. Funds should be prepared to justify their thesis, demonstrate discipline in underwriting, and highlight diversification strategies.

The founder perspective: navigating a shifting landscape

For founders, the warning signs are a double-edged sword. On one hand, capital is still available for truly differentiated ideas. On the other, the bar for funding has risen dramatically. Founders must focus on building sustainable businesses rather than chasing valuation arbitrage.

What investors want now

Founders should expect investors to prioritize the following:

  • Path to profitability. Even in the AI space, investors are increasingly focused on unit economics and the timeline to break-even. Founders should have a clear plan for achieving profitability within a reasonable timeframe.
  • Defensibility. AI startups must demonstrate a moat beyond just technology. This could include proprietary data, network effects, or regulatory advantages.
  • Real-world use cases. Investors are wary of companies that rely on hypothetical future demand. Founders should focus on solving tangible problems for identifiable customers.

The fundraising reality

The days of raising on a pitch deck alone are over. Founders must be prepared to provide detailed financial projections, customer traction, and a clear competitive landscape. The best-funded startups will be those that can show not just potential, but a clear path to monetization.

The LP view: what to watch in the coming quarters

For LPs, the current environment is a test of discipline. The temptation to chase hot sectors is strong, but the risks of overpaying for assets are equally high. LPs should focus on the following indicators to assess their exposure to AI-adjacent sectors:

Key metrics to monitor

  • IRR assumptions. Many funds are still using aggressive IRR assumptions for AI investments. LPs should scrutinize these assumptions and ask for sensitivity analyses.
  • Portfolio concentration. Funds with high exposure to AI-adjacent sectors may face significant markdowns if a correction occurs. LPs should assess whether their portfolios are adequately diversified.
  • Secondary market activity. A lack of liquidity in secondary markets could force LPs to hold onto underperforming assets for longer than anticipated. LPs should monitor secondary market trends closely.

Questions to ask your GPs

LPs should engage their general partners with tough questions:

  • How are you stress-testing your AI portfolio under a downside scenario?
  • What steps are you taking to protect against markdowns or write-downs?
  • Are you adjusting your underwriting standards for new investments in the space?

The bottom line: preparation over prediction

The warning from Cuban and Burry is not a prediction of doom, but a reminder that markets are cyclical. For funds, founders, and LPs, the key is to prepare for a correction without assuming one is inevitable. This means focusing on fundamentals, diversifying exposure, and building resilience into portfolios.

The dot-com bust was a painful lesson in the dangers of hype over substance. Today’s AI-adjacent sector may not face the same fate, but the risks are real. The difference between success and failure will come down to discipline, not luck.

What to do next: Reassess your exposure to AI-adjacent sectors and stress-test your portfolio for a potential correction.