The Hidden Economics Powering AI

Full Title

Summary

This episode examines how the economics of AI are rapidly evolving, with significant implications for private markets and venture capital. The discussion highlights the massive infrastructure build-out, accelerating AI capabilities, and the changing landscape of how companies stay private longer and how value is created and captured.

Key Points

The tech market is dominated by large US-based technology companies, and AI is accelerating this trend, with companies staying private longer than ever before.
Major tech companies are investing hundreds of billions in AI infrastructure, a build-out happening faster than previous technology cycles, which benefits companies building on top of it.
The cost of accessing frontier AI models has drastically decreased (over 99%), while model capabilities have doubled roughly every seven months, indicating rapidly improving input quality and cost-effectiveness.
AI is expected to be more transformative than the mobile/cloud cycle, with the potential to impact a much larger portion of the economy beyond just software spend.
A significant portion of the value created by AI will likely go to end customers, but companies that capture this opportunity can still achieve massive market cap growth.
The rapid adoption of AI is unprecedented, with tools like ChatGPT reaching massive user numbers significantly faster than previous internet giants like Google.
The supply side build-out for AI infrastructure is being funded more stably through banks and private debt, with insurance companies potentially involved, unlike some previous cycles.
Energy production is becoming a bottleneck for AI infrastructure, and nuclear power is seen as a promising solution, with companies building data centers near nuclear plants.
Cooling systems for data centers are emerging as a future bottleneck, driving innovation in that area.
The business model for AI companies is evolving, with a focus on customer retention, ease of acquisition, and flexible pricing strategies beyond traditional SaaS models, potentially including subscriptions and freemium with advertising.
Gross margins for AI companies are being scrutinized, but there's leniency due to the expectation of decreasing input costs from competition among model providers.
The public markets are no longer the primary venue for high-growth technology companies, with most residing in private markets, leading to extended periods of companies staying private.
Companies that can innovate in UI/UX, leverage new data sources, and implement business model innovations have the best chance of disrupting incumbents.
AI applications in specific areas like medical scribing, customer support, and high-end financial analysis are likely to be stickier due to integration and company-specific workflows.
The focus for venture capital is shifting towards companies with high-quality teams, even if business outcomes have a wider range of variance, and on identifying asymmetric bets with downside protection.

Conclusion

AI is fundamentally changing the economic landscape, creating massive infrastructure build-outs and accelerating innovation at an unprecedented pace.

Private markets are increasingly becoming the hub for high-growth technology companies, with AI driving significant value creation and new business models.

Investors need to adapt their strategies to understand the unique economics of AI, including infrastructure needs, evolving business models, and the rapid pace of adoption, while also considering future bottlenecks like energy.