The author further points out the paradox of modern AI, where companies pursue larger models and centralized data collection while promising personalized experiences. As AI improves at handling standardized decisions, the remaining "last mile" of human judgment becomes more valuable and harder to scale. The author suggests that tech companies could learn from the Catholic Church's balance of global scale and local knowledge. The article concludes by emphasizing the importance of finding the right balance between global AI systems and local expertise for optimal results.
Key takeaways:
- The principle of subsidiarity, which pushes decisions to the lowest effective level, can be applied to AI systems to better capture local expertise and human knowledge.
- AI systems often struggle with local context and specific circumstances due to their architecture prioritizing patterns across vast datasets over preserving specific, local knowledge.
- The paradox of modern AI is that as it gets better at handling standardized decisions, the remaining 'last mile' of human judgment becomes more valuable and difficult to scale.
- Companies that combine AI tools with local expertise have seen the highest ROI, suggesting a balanced approach between scale and local control is most effective.