Ash Fontana on AI, Investing, and Building Data Learning Effects

The Evolution of Technology & AI

• 💡 Technology provides leverage: physical (tools, weapons), intellectual (computers, calculators), and decision-making (AI, prediction).
• 🧠 The AI First Century began in the 1950s with neural networks, not recently, and is nearing its "pointy end."
• 🚀 Early AI research focused on understanding the nervous system and recreating neurons, leading to deep learning and transformers today.

Google's AI-First Approach

• ✅ Google was founded as an AI company from day one, leveraging PageRank as a machine-learnable algorithm.
• 📊 Their competitive advantage stems from distributed computing systems, acquiring vast datasets (e.g., Google Maps, Android), and developing learning algorithms.

Understanding Data Learning Effects (DLEs)

• 🔑 A Data Learning Effect combines a critical mass of data, a process to turn data into information, and an algorithm that learns and improves over time.
• 📈 DLEs are a powerful new source of competitive advantage, integrating elements of economies of scale, learning effects, and network effects.
• ⚠️ Companies without a DLE risk being subjugated to a system of record by new entrants with predictive systems.

Building AI-First Companies

• 🛠️ Founders should focus on identifying core predictions and using basic statistical methods with existing data, rather than immediately buying expensive AI tools.
• 🎯 To build a DLE, founders must define their unique data, the information it provides, a useful prediction, and a feedback loop for improvement.
• 💰 AI agents allow founders to achieve more with less, reducing the need for early hires and lowering startup costs, but increasing competition.

The Future of AI Integration

• 🌐 The next wave of AI involves deeper integration with daily life, moving beyond chatbots to real-world applications.
• 🤖 Future developments include brain-computer interfaces, neuromodulation, and autonomous robots that interact reliably in the physical world.