Similar to the dot-com era, the current AI investment cycle is expected to produce a high number of company failures alongside a few generational winners that create more value than ever before in venture capital history.

The most dangerous venture stage is the "breakout" middle ground ($500M-$2B valuations). This segment is flooded with capital, leading firms to write large checks into companies that may not have durable product-market fit. This creates a high risk of capital loss, as companies are capitalized as if they are already proven winners.

During the dot-com crash, application-layer companies like Pets.com went to zero, while infrastructure providers like Intel and Cisco survived. The lesson for AI investors is to focus on the underlying "picks and shovels"—compute, chips, and data centers—rather than consumer-facing apps that may become obsolete.

The dot-com era was not fueled by pure naivete. Many investors and professionals were fully aware that valuations were disconnected from reality. The prevailing strategy was to participate in the mania with the belief that they could sell to a "greater fool" before the inevitable bubble popped.

Investors often reject ideas in markets where previous companies failed, a bias they call "scar tissue." This creates an opportunity for founders who can identify a key change—like new AI technology or shifting consumer behavior—that makes a previously impossible idea now viable.

Success isn't linear. Mobile gaming giant Supercell didn't start with mobile games, and drone delivery firm ZipLine began with a robotic toy. This shows that foundational failures in one area can be the necessary learning experiences that lead to market-defining success in another.

Market dynamics are not static. What was once a 'wave'—a new, urgent problem for everyone—can evolve into a series of 'dams' and eventually a stable 'river.' A common mistake is to build for the hype of a wave after it has crested, by which point it no longer provides the same opportunity for explosive growth.

The dot-com era saw ~2,000 companies go public, but only a dozen survived meaningfully. The current AI wave will likely follow a similar pattern, with most companies failing or being acquired despite the hype. Founders should prepare for this reality by considering their exit strategy early.

When evaluating revolutionary ideas, traditional Total Addressable Market (TAM) analysis is useless. VCs should instead bet on founders with a "world-bending vision" capable of inducing a new market, not just capturing an existing one. Have the humility to admit you can't predict market size and instead back the visionary founder.