Some of our thoughts on investing in AI, at Quadratic VC.

Here's the TL/DR version.

AI companies can be categorised into three main types based on how deeply AI is integrated into their operations:

1. Point Solutions - These companies implement AI for specific, standalone tasks like fraud detection.
2. Application Solutions - These businesses integrate AI across multiple related functions, enhancing broader processes within existing frameworks.
3. System Solutions - These organisations fundamentally reimagine and reconstruct their entire business models around AI technologies.

Emerging Trends:

1. AI advances are reducing training costs, yet the overall cost of models is increasing exponentially.
2. The scalability of AI technologies has created significant competitive advantages, leading to a landscape where large firms hold substantial leverage due to their data resources.

Strategic Insights:

1. AI Scope-Innovation Matrix: Investors should focus on startups that are high on innovation across broad or narrow scopes. Such firms are likely to redefine markets and offer substantial returns, while low innovation firms offer limited growth potential.
2. Synthetic and Alternative Data: New entrants should consider leveraging synthetic data and alternative data sources to innovate and compete effectively in the AI space.


Predictions:

1. Open-Source AI Libraries Expansion: Firms like Meta will continue to open-source foundational AI models to boost innovation and reduce costs, a strategic move to enhance their core business ecosystems.
2. Strategic Hardware Differentiation: Companies such as Nvidia and Intel will intensify their focus on differentiating their AI hardware through proprietary technologies and exclusive partnerships.
3. Open Standards Rivalry: Software companies will advocate for open standards and open-source projects to challenge hardware-based differentiation.
4. Shift Towards Proprietary Datasets: The critical need for quality training data will drive companies and investors to strategically acquire data-rich companies, focusing on their data assets rather than just products.