Like containerization, AI is a transformative technology where value may accrue to customers and users, not the creators of the core infrastructure. The biggest fortunes from containerization were made by companies like Nike and Apple that leveraged global supply chains, not by investors in the container companies themselves.

AI companies defy old categories. They raise growth-stage capital while pre-revenue (like venture) and serve as both foundational platforms (infrastructure) and direct-to-user products (apps). This blurring of lines demands a new, hybrid approach from investors and founders.

Instead of betting on specific AI models like ChatGPT, a more robust strategy is to invest in the underlying infrastructure that all AI development requires. This 'onion' approach focuses on second-order essentials like semiconductors and data centers, which are poised to grow regardless of which consumer-facing application wins.

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.

When a new technology stack like AI emerges, the infrastructure layer (chips, networking) inflects first and has the most identifiable winners. Sacerdote argues the application and model layers are riskier and less predictable, similar to the early, chaotic days of internet search engines before Google's dominance.

To understand the crypto landscape, categorize assets by function. Bitcoin's primary role is a neutral, hard money store of value—like digital gold. Ethereum acts as a programmable settlement layer for stablecoins, tokenized assets, and AI agents—making it the system's digital oil.

AI and crypto are not competing but are parallel, complementary forces reshaping business. While AI revolutionizes company creation and internal operations, Internet Capital Markets (powered by crypto) are fundamentally rewriting the external functions of capital formation, trading, settlement, and ownership for this new generation of AI-native companies.

The enduring moat in the AI stack lies in what is hardest to replicate. Since building foundation models is significantly more difficult than building applications on top of them, the model layer is inherently more defensible and will naturally capture more value over time.

Rather than picking a winning AI or crypto, the smarter investment is in the 'picks and shovels.' This means focusing on the infrastructure every autonomous agent will require to transact—such as wallets, custody services, and blockchain rails—regardless of which specific application succeeds.