Gang Fang's Blog

A colleague once said that drones are not innovative because they are just four motors and some plastic put together, arguing that these are existing components. However, I don’t believe that this reductionist argument fully explains the true nature of these devices.

This particular argument stuck with me, as I had recently been contemplating a seemingly unrelated thought exercise regarding ChatGPT’s intelligence. I reached a conclusion that ChatGPT is essentially a stochastic carrot and can be reduced to something as simple as linear regression, as they are both based on the same philosophical idea of making predictions.

My understanding deepened when I read a blog post about the emergent abilities exhibited by GPT-4, which scales even more dramatically than its predecessor, GPT-3. This made me realize that there might be something unexplainable by reasoning solely from foundational and simpler ideas, but rather it could be attributed to the sheer scale of the system.

I also came across the concept of “Artificial Intuition,” which was used by Monica Anderson back in the 2000s. The idea that “intelligence must come from unintelligent mechanisms” made me recognize that this emergent phenomenon exists not only in AI but also in biology and various other human-scale systems, such as the economy, the internet, and stock markets.

It became apparent to me that emergence and complexity are more prevalent than I had previously realized. To understand complexity, the conventional reductionist approach in science may not always be sufficient. This approach involves breaking down a phenomenon into smaller, more comprehensible components. However, emergence brings qualitative changes to a system or its characteristics that cannot be fully explained through reductionism.

This alternative avenue of exploring complexity has opened up new frontiers for study. In fact, the Santa Fe Institute was established in the 1980s with a focus on the interdisciplinary study of complexity. It is crucial for me to be aware of this as I continue my own studies.

Currently, I am reading the book “Complexity: A Guided Tour” written by Melanie Mitchell. (https://www.goodreads.com/book/show/5597902-complexity).

Note: to be clear, the drone example isn’t an emergent phenomenon as they are a much simpler system, than let say the human immune system, that is designed based on well understood physical principles. But this example helped me connect the dot so it deserved a mention.