AI Investment Themes - Core AI (1/4)

Wayne Gretzky famously said, “I skate to where the puck is going, not where it has been.”

Introduction

At FuturProof, our guiding principle has always been to 'skate to where the puck is going, not where it has been.' Embracing this philosophy, I’ve dedicated myself to becoming a student of frontier markets, exploring opportunities that are often underappreciated or misunderstood.

I have spent the last year trying to understand AI and related technologies. My exploration has been twofold: as an investor seeking opportunities and as a builder looking to understand and leverage these technologies across our portfolio.

The critical question I explore is: What aspects of AI today represent groundbreaking shifts, and which are extensions of existing, long-term research?

In the coming months, I aim to share my insights and analyses of the AI domain. This series is designed not just to present what I've learned but to open a dialogue. Together, we can enhance our collective understanding, challenge assumptions, and uncover the true potential of AI in shaping our future.

Series Format

In my analysis, I categorize the main AI investment themes into four distinct yet interconnected categories:

1. Core AI: This category encompasses the foundational elements of artificial intelligence. It includes the development of fundamental AI technologies, such as advanced algorithms, underlying research, and the base models that form the backbone of AI capabilities. This theme is where groundbreaking AI theories are translated into practical models, shaping the future trajectory of AI development.
2. Applied AI: This theme focuses on the real-world implementation of AI technologies across a diverse array of industries and applications. From healthcare diagnostics to financial forecasting, and from enhancing retail experiences to optimizing manufacturing processes, Applied AI represents the direct utilization of AI advancements to solve specific, practical problems and improve efficiencies in various sectors.
3. AI Hardware: Essential to the functionality of AI systems, this theme includes the physical components that power AI processing. It involves investments in cutting-edge hardware such as GPUs, specialized AI chips, and other innovative hardware solutions designed to meet the intensive computational demands of advanced AI operations.
4. AI Infrastructure: Underpinning the functionality of AI systems, this category encompasses the critical backend technologies and platforms. It includes investments in vector databases, cloud computing platforms, and other essential infrastructure components that support the seamless operation, data management, and scalability of AI systems. This infrastructure serves as the foundational framework enabling AI applications to function efficiently and effectively.

Core AI: Details

I will focus on Core AI for the first edition of this series. Core AI is a broad term that includes various subfields and components such as:

1. Fundamental Theories and Concepts: Core AI begins with the theoretical basis of AI, including the principles of machine learning, neural networks, deep learning, reinforcement learning, and other computational theories that enable machines to perform tasks that would typically require human intelligence. This includes understanding human cognition and decision-making processes and translating these into computational models.
2. Algorithms and Models: At the heart of Core AI are the algorithms and models used to process data and learn from it. This includes supervised, unsupervised, and semi-supervised learning algorithms, along with more specific approaches like decision trees, support vector machines, neural networks, and deep learning architectures. These algorithms are what enable AI systems to recognize patterns, make predictions, and make decisions based on data inputs.
3. Natural Language Processing (NLP): A critical component of Core AI, NLP involves the development of algorithms that allow computers to understand, interpret, and generate human language. This includes tasks such as speech recognition, language translation, sentiment analysis, and text generation.
4. Computer Vision: This aspect of Core AI is concerned with enabling machines to interpret and understand visual information from the world, such as images and videos. It involves the development of algorithms for image recognition, object detection, pattern recognition, and image generation.
5. Machine Perception: Core AI also includes the development of systems that can perceive and interpret sensory data, mimicking human senses. This goes beyond just vision to include hearing, touch, and potentially other senses, allowing AI systems to interact more naturally and effectively with their environment.
6. Ethical AI and Bias Mitigation: As AI technology advances, Core AI increasingly encompasses the study and implementation of ethical AI practices. This includes understanding and addressing biases in AI algorithms, ensuring transparency, and considering the ethical implications of AI systems.
7. AI Frameworks and Tools: Core AI also involves the development and improvement of frameworks and tools used to build and deploy AI models. This includes libraries and frameworks like TensorFlow, PyTorch, and Keras.

In essence, Core AI is about building the fundamental capabilities of AI systems, enabling them to learn, adapt, make decisions, and perform tasks that would otherwise require human intelligence. It's a dynamic field that continues to evolve, driven by both theoretical research and practical advancements.

Advantages

1. Foundational Impact: Investments in Core AI have the potential to influence multiple industries, as these technologies underpin various AI applications across sectors.
2. Long-Term Growth Potential: The expanding AI field suggests that Core AI technologies could offer significant growth and value in the long term.
3. Broad Exposure: Investing in Core AI provides exposure to a wide array of emerging technologies and applications, instead of picking specific applications of the technology.
4. Intellectual Property and Leadership: Investments can lead to the creation of valuable intellectual property and establish leadership, enhancing competitive advantage at both the model and application layers.

Disadvantages

1. High Research and Development Costs: Developing foundational AI technologies is resource-intensive and expensive, with extended periods before seeing profitable returns.
2. Technical Complexity and Uncertainty: The field's rapid evolution and technical complexity add layers of uncertainty and risk to investments.
3. Regulatory and Ethical Challenges: Advancing AI technology faces increasing scrutiny regarding ethics, privacy, and regulation.
4. Difficult Monetization: It can be challenging to directly monetize Core AI innovations, as the path from research to commercial product is often long and uncertain.
5. Competition from Tech Giants: Large tech companies with substantial resources, distribution, and data can dominate the Core AI field, making it difficult for smaller entities to compete.
6. Competition from Open Source: Open-source AI projects and frameworks can dilute the market, making it harder for proprietary solutions to become commercially sustainable.
7. Dependence on Continuous Innovation: Core AI requires ongoing innovation to stay relevant, necessitating a constant influx of resources and novel ideas.

Compared to investing in other AI themes, investing in Core AI is more about betting on foundational research and long-term potential rather than immediate practical applications or specific market needs. This fundamental difference shapes the risk-reward profile of investments in Core AI.