Yesterday, I visited an Exhibition on contemporary marketing in Tokyo Big Site with my colleague in charge. Actually, we didn’t want to see the exhibition per se, but take part in a special session where some Japanese experts of artificial intelligence (AI) delivered speeches on its current developments.
To be honest, it was extremely boring and disappointing indeed. Based on their professional firm belief that the “singularity” will come true within decades, they simply tried to persuade us to adapt ourselves to the “new reality” with AI. According to their explanations, we won’t survive without accelerated technological development of AI, and only what is left for us to accept this challenge without any objections and go along with AI, which is said to be becoming much clever than us.
Well, that sounds really good, guys. Nevertheless, my smart colleague HA and I left the room even before the ending of their speeches, because they pointed out one thing that I could hardly understand: One of the experts who’s actually the “star” in this sector, told a story in terms of how to make AI write a novel. During the experiment, he gave a series of texts to AI, and AI decided to line them up randomly, as if that had been written by a human author. The expert praised the experiment was so successful that every human author will be replaced by AI sooner or later.
Spontaneously, it occurred to me that this guy is an expert of not “art” but just “algorithm”. He’s not capable of explaining how to generate original “texts”, with which AI makes something like “novel”. As far as I correctly understand, it’s he as a human being that gives commandos to AI to line up his texts randomly. That’s never a proof to show us AI will be soon given its own creativity. As you may know, I’m author and have been writing thousands of sentences and texts. I’m wondering how I’m used to practice to “write”: I’m always given an image in my brain and try to delineate what it is. That’s it and exactly what I regard as “creativity”. Creativity means generating something from nothing. It’s neither static analysis nor algorithm.
On the way back from Tokyo Big site to my office located in Sengokuyama/Tokyo by taxi, I remembered the famous discourse Japanese leading thinkers had in the pre-wartime. It was called “discourse on overcoming the modern （近代の超克）”. From time to time, I got back to what they discussed in 1930s and 1940s in this regard and stuck myself to develop the results of that. So far, I couldn’t decide to which direction I should move forward, however, I’m feeling I’ve finally got it.
The discussion on “singularity” AI is said to bring about and the reality we’ve been experiencing in the global community clearly show the dichotomy we, all the human beings are commonly facing. AI never votes for “Brexit”, because it’s not rational and against algorithm. But the reality is the people on the street in UK voted for self-exclusion from EU. Why? Because we, the human beings consist of both rationality and irrationality. Even the western philosophers in the modern time were quite aware of that and made effort to build up “philosophy of the life (Lebensphilosophie (F. Nietzsche)”. “Menschliches, allzu Menschliches” was their motto and reaction to exaggerated rationalism getting along with industrial revolutions. The latter is usually called modernism, which still controls the way of our thinking.
Proven by “Godel’s incompleteness theorems”, singularity will never come true. All the experts of AI know that, but pretend as if they can overcome the theory. What we should overcome is rather the naïve modernism symbolized by firm belief in technological development initiated by simple rationalism. All we need is therefore “overcoming the modern” as such. The decision for “Brexit” and other political phenomena we see nowadays (D. Trump’s rising and “AfD movement” in Germany, for example) are omens to show the dawn of the new human era. What we need is not algorithm but deep thought on seeking the reason why we were born, (have to) live and pass away. Such a single fact can’t be figured out by AI. Never ever.