The author suggests that creating a causal model is the first step towards achieving "strong" AI. This involves identifying a foundational process critical to the organization's success, capturing the true logic running operations, and recognizing marginalizing effects in the longer term. The article concludes by stating that while AI influenced by causal modeling is projected to be a few years away, some companies are already well down this path.
Key takeaways:
- For AI to evolve from being weak to strong, it needs to be coupled with more sophisticated causal and other modeling, allowing solutions and recommendations to become more sustainable and robust.
- Empirical data cannot anticipate edge cases or adequately prepare an organization to respond to changing markets. To achieve a more autonomous AI, models should embed non-empirical layers such as mathematical-based rules and insights and meta-knowledge.
- Creating a causal model is crucial to capture the true logic running organizational operations, allowing AI to make a more positive long-term impact and avoid unintended consequences.
- Organizational life has a high degree of complexity embedded in it and causal-based AI systems will contribute to more effectively managing this complexity.