Ant Society and AI: The Phenomenal Information Processing Capabilities of the Natural World

With the recent evolution of Natural Language Processing (NLP) and Large Language Models (LLMs), when examining the theories of information engineers, neuroscientists, philosophers, and others, if technological innovation continues at this pace, it can be broadly categorized into three views: (1) AI will become smarter than humans, (2) it is impossible for AI to become smarter than humans, (3) it is unclear what will happen.

Based on that, (4) there is also the perspective that the definition of what is 'smart' is not clear, but what do you, the readers, think about the evolution of AI?

The reaction the author expects from readers is a 'Don't ask ordinary people about things that even the world's most brilliant scientists can't agree on!' retort, but if readers ask themselves 'Are humans truly smart?', they might find this kind of discussion enjoyable. Thinking about the

'difference between a genius and an ordinary person' is also an interesting theme. Most people probably think of a genius as someone with a level of intelligence in a different dimension from an ordinary person. People who speak or think about things we can understand are, after all, just ordinary people and do not meet the author's definition of a genius.

The 'ABC conjecture', known as a super-difficult problem in the mathematical world, is a mathematical conjecture proposed by Dr. Joseph Oesterlé and Professor David Masser in 1985, and is also called the 'Oesterlé-Masser conjecture' after the two who proposed it.

'ABC conjecture' will be called the 'ABC theorem' once it is mathematically proven, but the peer review of the paper written by Professor Shinichi Mochizuki of Kyoto University took seven and a half years. In this case, Oesterlé and Masser, who proposed the ABC conjecture, and Professor Mochizuki, who attempted to prove the problem with his own unique approach, can be considered geniuses. On the other hand, if the ABC conjecture itself is proven wrong, all three might be called fools. This is the reason for the saying, 'The line between a fool and a genius is thin'.

This problem, which is difficult for even all the mathematicians in the world to prove or disprove, is beyond the understanding of ordinary mathematicians, so it can be said that this is truly the world of geniuses.

However, isn't the difference between a genius and an ordinary person far smaller than the difference in intelligence between a monkey and a human? This also shows just how significant our preconceptions and biases regarding human intelligence are.

From the perspective of an information engineer, I am amazed by the capabilities of insects, especially ants and bees. For example, even an ant only about 1 millimeter long performs numerous complex tasks simultaneously with its tiny body.

Digestion and absorption of food, control of leg movements, identification of the colony based on subtle differences in formic acid, searching for and transporting food, and even sanitation management within the colony, care for the queen ant, and management of eggs and larvae—these actions are carried out smoothly within the simple yet complexly integrated society of the ants.

1-2 millimeter-sized ant robots have already been developed, but they are still far from reaching the capabilities of biological ants. The development of microrobots three orders of magnitude smaller than 1 millimeter, and nanorobots three orders of magnitude smaller than that, is also progressing, and the cutting-edge technology in this field is as interesting as the cutting-edge field of AI.

In the world of organisms like colony-forming ants and bees, intelligence is studied under the concepts of collective intelligence or collective intellect, and research into collective intelligence is also progressing in the world of computer science.

When controlling the behavior of these organisms artificially, it is a difficult challenge for an information engineer to mimic and control the behavior of a single ant with an independent single-chip CPU. This fact clearly demonstrates the gap between the phenomenal information processing capabilities of natural organisms and the technology that attempts to mimic them.

Reflections on the functions and intelligence of the human brain provide a good opportunity to consider how superior, or conversely, how inferior AI is compared to the natural world and other organisms. And as a result, it will likely serve as a trigger to rethink philosophical questions about what humans are, and what human existence and free will are.

Paradoxically, problems that could not be solved by previous philosophical approaches might find clues to their solutions by thinking through AI.