Infrastructure

Contrary to popular belief, the best time to invest in infrastructure is after applications have been successfully adopted. Amazon built AWS (infra) after video streaming, e-commerce, and web applications attracted millions of users. Similarly, electricity (infra) became widely available only after lightbulbs (app) were invented, and airports (infra) were built after airplanes (app).

Therefore, it's not the best use of capital to invest in the infrastructure layer between AI models and applications. Instead, investing in the infrastructure needed for LLMs themselves, such as new computing hardware, synthetic data for training, and tools for efficient training will lay the groundwork for future innovation and adoption.

Applications

From a market perspective, there are four categories of AI products :

1. Enhancing existing products and workflows by incorporating AI capabilities with the core value props of improving efficiency, accuracy, or user experience in existing processes. Examples include:
   • Productivity tools like information reterival and agents for knowledge workers like lawyers, accountants, investment analyst, scientific researchers, supply chain procurement, electronic engineering, and business consultants. Core benefits are automation, faster access to information, and better user experience.
   • Content and text generation platforms that streamlines workflows like HR training and coaching platforms powered by language models and generative AI and end-to-end marketing platform from brainstorming to content creation.
   • Data transformation, generation, and enrichment products that structures data into different formats and creates synthetic data. These services can significantly improve business processes such as healthcare payment systems and freight forwarding and trade documentation.
2. New products or services that leverage AI to create new possibilities by lowering the learning curve of existing tools, empowering non-consumption and creating new markets. Examples include generative architecture and interior design tools, legal document creators, and form fillers for a variety of use cases and verticals.
3. Disrupting existing products and markets by offering AI-powered alternatives at significantly lower costs. AI replacing outsourcing services in (a) accounting, financial, and insurance, (b) procurement and customer service, and (c) IT and consulting are ripe to be automated with AI.
4. Products that cater to specific niches within existing markets with tailored AI solutions. Examples include:
   • Business in-a-box solutions powered by AI solving all major problems for nascent segments (e.g., AirBnB and Turo hosts, influencers on social media, and YouTubers) and/or existing ones (e.g., SMBs, and loan brokers) with integrated software solutions.
   • AI-enabled platforms that create new opportunities for entrepreneurs and businesses to earn money using or building upon the platform, creating new value networks. These platforms solve key technology and business problems with a simple solution and ****enjoy competition among their users and complimenters such as Figma, AirBnb, and Shopify.

Frontier opportunities

1. Frontier AI technologies, including symbolic and causal reasoning, complex and hierarchical planning, sequences and subtasks decomposition, short-term memory networks for AI, dynamic knowledge graphs and ontologies, explainable AI systems, adaptive real-time AI models, probabilistic programming frameworks, AI control systems, ground truth platforms.
2. Neural networks and foundational models for science. Some major areas include chemistry (e.g., molecule discovery), physics (e.g., solving Maxwell's equations), biology (e.g., gene analysis, protein folding, and drug discovery), electronics (e.g., analog chip design), and math (e.g., solving NP-complete problems).
3. Horizontal AI platforms for new markets such as bioinformatics, manufacturing digitization, and robotics. These platforms foster the growth of the whole sector and, from an investment standpoint, act like an ETF for that vertical.