The Aesthetics of Ambiguity - Fuzzy Control and AI

Is AI neutral? And are we truly living by 'choice'?

By contrasting the once-prominent 'fuzzy control' with today's generative AI, we examine the 'loss of inquiry' behind convenience. This is a small clue for thinking about technology, control, and humanity.
This is a small clue for thinking about technology, control, and humanity.

-- Humanity as considered through the contrast between fuzzy control and AI

Aren't you just thinking that you're amazing for saying 'ChatGPT is amazing!'?
There was a time when I thought that way too.

This note is the 'Era of Intellectual Polarization - Those Who Can Manipulate AI and Those Who Are Manipulated' (7 parts total) first installment.

Vol.1 The Aesthetics of Ambiguity - Fuzzy Control and AI
Vol.2 The “Joy of Learning” Stolen in the AI Era
Vol.3 The Intelligence of Discomfort in the AI Era
Vol.4 Invisible Guidance - The “Obedient Society” Optimized by AI
Vol.5 Who Monitors the AI? - In Search of Technological Neutrality
Vol.6 Nuclear Power and AI - Technology That “Won't Let You Quit” Through Fear and Power
Vol.7 A New Democracy in the AI Era - The Potential of a Community That Asks Questions

Introduction: When Those “Fuzzy Appliances” Were All the Rage

Do you remember the term “fuzzy control”?

From the 1980s through the 90s, commercials for home appliances like air conditioners and microwave ovens would repeat the word “fuzzy” over and over. These “smart appliances,” which incorporated human sensibilities to automatically provide comfort, were treated as the cutting edge of technology.
However, after a certain point, that word suddenly stopped being heard. It wasn't the technology itself that disappeared, but rather the excessive expectations that had vanished.

I have had a similar experience. In the era of PC communications, there was a program called “Artificial Incompetence.” It was a so-called chatbot, a parrot-like entity that would respond to a user's words with appropriate-sounding nods. Conversations didn't really happen, but some people found it amusing. It was a game. It wasn't technology, but rather an extension of “make-believe play”.

That is why, when AI became a hot topic again, I was somewhat indifferent.
“It's just another fad like that,” I thought.
“Isn't this just a rehash of fuzzy control or artificial incompetence?”

Certainly, I was surprised when the world chess champion lost to a computer. But I didn't think it would change my own life. Even when I heard that AI was trading on the stock exchange, it felt like something far removed from my reality.

However, since the arrival of ChatGPT, something has changed decisively.

It doesn't just respond to words; it understands context, supplements intent, and even picks up on what is left unsaid.
Moreover, that ability is evolving rapidly, month by month.
What started out unable to perform simple calculations is now said to be capable of passing the bar exam.

Amidst all this, I watched footage on television of the silent New York Stock Exchange.
In that place where loud voices once flew and paper fluttered, AI now quietly handles the trading.
There was no longer any room for human judgment to intervene.

In Shogi, too, AI made an opening move that “a human would never choose,” and it won.
When asked for the reason, the developer replied, “I don't know.”
We have entered an era where an intelligence we cannot understand shows us only the results and wins. era.

Even so, we must ask.

• Is AI truly neutral?

• Does it contain no one's intentions?

• How do we live in a society that follows its judgments?

Chapter 1: Is AI a "Neutral Third Party"?

AI is often thought of as a "third party" that makes fair and rational judgments.
We want to believe that it chooses the best option calmly, without emotion or personal bias.

However, AI is neither a god nor an angel. It is an artifact created by humans.
The data it learns from and the values it is designed with are all provided by humans.

For example, the American judicial support AI "COMPAS" gave unfairly high scores to Black defendants when assessing recidivism risk.
The cause was that the past judicial data used for training already contained racial bias.

A company's recruitment support AI learned the characteristics of previously hired employees and unconsciously rated women and older people lower.
The AI did not learn the pattern of "excellent talent."It merely reproduced the biases of human society as they are.

Even so, we want to believe it because "the AI says so."
We want to believe that because it is an AI without emotion or personal bias, it makes fair judgments.
But in reality, behind its output, there is always someone's judgment, someone's design, someone's intention.
And that "someone" is usually invisible to our eyes.

The responses of generative AI are also not completely free.
For certain themes or positions, there is a pre-set "filter."
Even expressions that seem neutral at first glance have the discretion of the operators working behind them regarding "how much to say / what to avoid."

In other words, AI is like a black crystal that looks transparent but is opaque.
"Who designed that judgment?"
"For whom is that answer convenient?"
If we do not keep asking, we may stop thinking amidst a "convenient correctness."

Chapter 2: Intelligence Inside the Black Box

When a Shogi AI defeated a professional player with an "impossible move,"
the developers could not explain the meaning of that move.

"We don't even know why that move is strong."

I was deeply shocked by these words.
AI has now stepped into a realm that cannot be understood by human logic or intuition.

Through neural network and deep learning technologies, AI decisions are now derived within a "sea of numbers."
The process is too complex for human language or thought to keep up with.
Only the correct answer is shown, and the reason remains unknown.

The same thing is happening in the world of finance.

Now, many trades on the New York Stock Exchange are executed by AI in microseconds.
There is almost no room left for human intervention.
Why did stock prices surge? Why did they fall?
There is no longer anyone who truly understands those movements.

The foundation of the economy is governed by logic that humans cannot understand.
Even so, we buy goods and live our lives at prices determined by AI.

It is as if an invisible intelligence is secretly designing society. itself.

Chapter 3: The Value of Mediocre Knowledge is Vanishing

Previously, "people who could do a little bit of everything" were valued.
Being able to write a little, code a little, or read a little English.
Such skills were reliable weapons in society.

But now, they have become areas that AI takes over in an instant.

Formatting reports, summarizing documents, and structuring presentations—AI handles them all instantly, without fatigue, and as many times as needed.

A world where skills acquired through hard work can be reproduced with the push of a button.
Where should we find the 'meaning of learning'?

Even so, there are things AI cannot do.

That is to pose questions.

AI is an excellent 'answerer,' but humans must decide what should be asked in the first place.
For whom, for what purpose, and toward what kind of future.

To be a being that uses AI rather than being used by it,
one must not be a 'person who knows the correct answer,' but a 'person who can continue to hold questions'.

Chapter 4: Will those who can manipulate AI become the new 'rulers'?

The number of people who 'use' AI in their daily lives is increasing rapidly.
However, those who can understand, manipulate, and design its 'inner workings' are very few.

The output of generative AI appears neutral and inorganic at first glance.
However, what determines its behavior are the algorithm designers, the selectors of training data, and sometimes nations or giant corporations.

What if only a few experts could access the contents of AI, while everyone else only received the 'results'?

That would be the birth of a new privileged class that controls social rules through technology.
They hold immense responsibility and influence, like doctors or judges.
But the difference is that their existence is invisible.

As AI blends into daily life and we become less aware of its judgments, the power of the designers quietly grows stronger.

The asymmetry of knowledge is expanding, and a dichotomy is emerging between those who control AI and those who are optimized by AI. The middle ground is only fading away.
The middle class is now just fading away.

Conclusion: To continue being someone who asks questions

AI has made our lives more convenient.
However, are we truly trying to see what lies behind that convenience?

Fuzzy control and artificial stupidity were both hyped and then faded away.
That is why I was skeptical when ChatGPT became a hot topic.
But now, its evolution is completely different from the past in both speed and quality.

This time, something is truly beginning to change.

That is precisely why we must continue to ask not only "what can be done," but also
"is that really okay?".

It may be difficult to understand completely.
But anyone can ask questions.
And those questions are the final intelligence that prevents us from ending up as mere beings to be manipulated.

★☆★ Thank you for reading until the end.
If you resonated with this, I would be happy if you could press the "Like" button. Feel free to leave a
comment as well. The words you share will become hints for my next reflections. If you
follow me, I can deliver future articles to you.

Here are some recommended articles. I would be happy if you could read them.

▼ A society where AI is rapidly spreading. I explored how work styles and the value of skills will change.

▼ In an era where the areas where humans cannot beat AI continue to expand, what is it that only humans can do?

▼ This is a magazine that compiles this article series.