Could AI fully replace humans?

Day 23 / 365

When I first read about Polanyi’s Paradox in the book “Building a Second Brain,” I couldn’t help but be intrigued by its implications for the future of artificial intelligence and its relationship with human beings. Polanyi’s Paradox, first introduced in the book “The Tacit Dimension,” argues that there is a specific type of knowledge that humans possess that is difficult to verbalize or explain but is easily demonstrated through actions and experience.

One example of this is our ability to recognize faces. We are able to do this without being able to fully explain how we do it. This type of knowledge, known as tacit knowledge, is a fundamental aspect of human reasoning and understanding and is currently a major obstacle for AI in its attempts to replicate human capabilities.

This presents a problem for AI, as it is trained in a result-based way rather than a rule-based way. It relies on explicit knowledge, or raw data, to understand and complete tasks, but this needs to account for the tacit knowledge that humans possess. This is why we continue to see limitations in AI’s adaptability and interpretability, such as self-driving cars being unable to make unexpected detours or cleaning robots being vulnerable to unmonitored pets or children.

However, with recent developments in AI, such as ChatGPT, we are seeing a shift towards machines being able to infer tacit knowledge from examples and context without human assistance. This is a promising development, as the goal of having machines perform tasks is to replicate our outcomes for practical purposes, rather than our means.

In conclusion, Polanyi’s Paradox is a fascinating concept that highlights the unique capabilities of human beings and the challenges that AI faces in replicating them. While it may be concerning for some to think about the implications of AI replacing human jobs, it’s important to remember that the goal is to augment human capabilities and make our lives easier, not to replace them entirely.