As a professional in your field, you’re probably familiar with the many advancements that have been made in the field of artificial intelligence (AI) in recent years. AI has become more powerful and sophisticated, capable of doing many things that were once thought to be impossible. However, despite all these advancements, there are still some areas where AI falls short, and one of those areas is Level 3 reasoning.
To understand why AI is unable to perform Level 3 reasoning, we need to explore what that term means. Level 3 reasoning is the ability to understand causation, correlation, and counterfactuals.

Once there was a scientist named Dr. Jane who was working on a project to develop an AI system that could predict the likelihood of a person developing a certain disease based on their lifestyle habits. She was using a large dataset that included information about thousands of people’s lifestyles, medical histories, and other relevant factors.
Dr. Jane began by training the AI system to identify correlations between different variables in the dataset, such as the correlation between smoking and lung cancer. The AI system was able to do this quickly and accurately, identifying many correlations that had not been previously discovered.
However, Dr. Jane soon realized that correlation alone was not enough to predict whether a person was likely to develop a certain disease. She needed to understand the underlying mechanisms behind these correlations, to determine whether one variable was causing another. For example, was smoking causing lung cancer, or was there some other factor that was causing both smoking and lung cancer?
Dr. Jane began to explore the possibility of using the AI system to understand causation. She fed the system more data, hoping that it would be able to identify the underlying mechanisms behind the correlations. However, she soon realized that the AI system was unable to do this.
The AI system could identify correlations between variables, but it did not have the ability to understand the underlying mechanisms behind those correlations. It could not identify causation, and as a result, it could not accurately predict the likelihood of a person developing a certain disease based on their lifestyle habits.
Dr. Jane then turned her attention to counterfactual reasoning, which is the ability to reason about what could have happened if a certain event had not occurred. For example, what would have happened if a person had not smoked? However, she soon realized that the AI system was also unable to do this.
The AI system was unable to simulate alternate realities or understand how changes to one variable affected the outcome of a system. As a result, it was unable to accurately predict the likelihood of a person developing a certain disease based on their lifestyle habits.
Dr. Jane was disappointed that the AI system was unable to perform Level 3 reasoning, but she was not deterred. She continued to work on the project, exploring new ways to address the limitations of AI.
In conclusion, while AI has made many impressive advancements in recent years, it is still unable to perform Level 3 reasoning, which involves understanding causation, correlation, and counterfactuals. While AI can identify correlations between variables, it lacks the ability to understand the underlying mechanisms behind those correlations or simulate alternate realities to understand the impact of changes to one variable. As AI continues to evolve, researchers will find new ways to address these limitations if AI is going to replace humans who possess Level 3 reasoning, which I believe is not going to happen for a great reason. We are Humans after all.