Despite tech industry hype, mainstream AI adoption is incredibly shallow globally. This indicates the current boom is just the beginning and that the ecosystem is still very early. The contrarian risk is that the slow, massive, long-term diffusion of AI is being underestimated, not that the current moment is a bubble.
In a world of commoditizing technology, the most durable competitive advantage is trust. Building a core team from a founder's deep personal network—like former roommates and colleagues—creates a high-trust unit that can execute with extreme speed and flexibility across different domains, forming a powerful moat.
AI companies run private compute clusters at low utilization, similar to early industrial factories each having their own inefficient steam generator. This creates massive waste. The solution is a shared, coordinated compute grid that acts as an independent system operator to drive up utilization across the ecosystem.
The real synergy in a GameStop-eBay merger is using GameStop's 1,600 retail locations to physically verify high-value collectibles sold online. This verification layer is crucial for AI agents to confidently transact on behalf of users, solving the biggest hurdle in used-asset commerce: fraud.
Choosing a Public Benefit Corporation (PBC) structure is a strategic legal defense. It shields a company from shareholder lawsuits when making decisions—like providing compute at cost—that prioritize long-term ecosystem value over short-term profits, protecting the firm's core mission.
Top VCs are reviving the early, hands-on model of pioneers like Arthur Rock. Instead of just investing, firms are co-designing new labs from scratch, providing compute, capital, and commercial guidance. This "company creation" approach is viable again as capital is no longer the primary bottleneck for ambitious, frontier-tech ideas.