Brian Yeo, Chief Investment Officer of Singapore's sovereign wealth fund GIC, has presented a fresh perspective on the dynamics of the artificial intelligence market. In his view, the widespread adoption of open-source Chinese models will act as a catalyst for fundamental changes in the industry. The main effect will be intensified competition and, consequently, a significant reduction in the cost of implementing AI solutions for businesses.
Yeo emphasizes that technology accessibility is a key factor that will accelerate its mass integration into corporate processes. Cheaper and more open alternatives will allow companies not only to cut costs but also to significantly expand the range of AI use cases. This is especially important for segments where implementation was previously unprofitable due to high development and licensing costs for proprietary models.
GIC's Position: Betting on Infrastructure and Caution
The fund itself has already made a significant bet on the AI sector, investing billions of dollars in related assets. A prime example is GIC's leadership in the February funding round for Anthropic, totaling $30 billion. The fund expects further growth in the number of Chinese AI projects, but its approach to valuing startups remains measured. According to GIC analysts, the main risk is the disproportionately high cost of developing foundational models, which may not pay off for many young players.
My comment: Yeo's thesis about the declining cost of implementation is absolutely correct, but one should not forget the dual effect. On one hand, cheap open-source models democratize access to AI. On the other, they create risks for investors who have put money into "heavy" proprietary solutions. The market will likely see consolidation: either giants with unique data or niche startups with highly efficient models will survive.