Questioning the Machine

On ChatGPT, the I Ching, and encountering non-human Intelligence

Asking questions could be one of the unique factors that makes us human. Perhaps our tendency to ask questions is innate to our language. This questioning characterizes our inner drive to search for answers.

It is then no wonder that the “search engine”, revolutionized by Google, is what opened up the internet for the vast majority of people across the globe. As a recent article in the Economist puts it, “[f]or more than 25 years, search engines have been the internet’s front door” which has been dominated by Google ever since.¹

However, there is now a new challenger to the dominance of the Google search engine - the Artificial Intelligence (AI) Chatbot. Ever since its “re-release” in January, ChatGPT3 has started a new frenzy on AI discussions spawning not only media dialogue but a virtual AI war among tech platforms, specifically between Microsoft, which is partnering with ChatGPT3 to integrated with its own applications including Bing and Edge and Google which just announced its own AI chatbot, Bard.

ChatGPT3 is an AI chat box that allows you to ask it text based questions and receive real time responses. Many have logged in and asked it questions posting these answers on various platforms. This begs the question, what do you ask an AI? And what type of answers would you expect to receive?

I’ve been thinking about ChatGPT since this release and what this next gen AI could mean for us. I was driving my son to Jui Jitsu class recently when he said out of the blue, “You know how Einstein defined a genius? He said that the way to define creative genius is someone who is able to find multiple ways to solve the same problem.”

I thought about that for a second, finding many different ways to solve the same problem. Isn’t the whole point to solve the problem? Or maybe keeping the problem open to find another solution is more important than an answer?

One person who questioned existing systems of thought and language was Gottfried W. Leibniz. Leibniz was a 17th century philosopher who was a polymath and one of the inventors, along with Newton, of Calculus. One of Leibniz’s projects was trying to create a simpler number system for calculation than the 10 digit system inherited from the Indo-Arabic world, ie. 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. In this decimal system, which is the predominant number system for us today, the number sequence repeats every 10 digits as the numbers progress.

Instead of 10 digits, Leibniz created a number system with just 2 digits, a binary system of just 1 and 0. He lays out how to write numbers using the binary system below.

While the sequence is larger than the decimal system, Leibniz argues that this system makes calculations much easier because there is no need to memorize all the different arithmetic and multiplication tables. You can read his explanation of the binary system here.

With this binary system, operations follow a certain pattern. Discovering these recurring cycles can be used to find numbers without having to calculate numbers in your head, you just apply the pattern to find the number. Leibniz states,

And it even turns out that square numbers, cubic numbers, and other powers, likewise triangular numbers, pyramidal numbers, and other figure numbers, have similar cycles, so that tables of them can be written immediately, without any calculation. And this one drawn-out task in the beginning, which then gives the means to make reckoning economical and to proceed to infinity by rule, is infinitely advantageous.

Leibniz lays this out in his article “Explanation of the Binary Arithmetic” which was the foundation of the binary system which would later become the basis of the digital computer. When computers were first programmed they used the binary code for their language and it is still the predominant language used for computers.

What is interesting to note is that Leibniz was influenced by the ancient Chinese divinatory manual, the I Ching, translated as the Book of Changes. In his article he states how the drawing of the Trigrams, the basic building block of the I Ching, follow this same binary logic. He states,

What is amazing in this reckoning is that this arithmetic by 0 and 1 is found to contain the mystery of the lines of an ancient King and philosopher named Fuxi, who is believed to have lived more than 4000 years ago, and whom the Chinese regard as the founder of their empire and their sciences.

Here is the code Leibniz adds in his article to link the I Ching to his new binary system. The I Ching trigrams are drawn on the left column for the first 8 digits.

The trigrams of the I Ching are based on the divination of yarrow stalks which are used as part of the oracle. Like any oracle, the I Ching was believed to be a consultation with some ‘higher’ intelligence where a question was asked and one got an answer. The answer was initially given as a “yes” and “no”. The unbroken line (—) was a “yes” and the broken line (- -) was a “no”.

Over time this evolved into the 8 trigrams of three lines which can be seen in the left column in Leibniz’s table above and also below in a more modern rendering of the trigrams. The table below shows the 8 trigrams along with each name and image.

The 8 images are the 8 possibilities of three lines using this binary logic of broken and unbroken lines. Leibniz used this set of 8 to code the first 8 numerals (0 - 7) and then this pattern repeats by adding another column to the left and starting over at 8.

As we mention above, this binary code would become the basis of computer language that would spawn the digital age. The 1s and 0s of the digital code are each called a bit. What is interesting to note is that the smallest unit of the digital code is called a “byte”. A “byte” is composed of 8 bits and is the smallest unit to represent a character in a computer. Starting from bytes we grow to kilobytes (kb), megabytes (mb) and onto Gigabytes (GB) and Terabytes (TB) which we are now familiar with as units of storage on our devices. But all of this goes back to the arrangement of 1s and 0s in groups of 8, just like the I Ching.

While the computer age brought the computing power to process these groups of 8 bits to enormously large numbers, the I Ching only used a second level of the 8 images. From this second second level of 8 images emerges 64 different combinations of hexagrams (two different trigrams) which is the basis of the Book of Changes. The combination occurs with one trigram on top and the other on the bottom and generates a corresponding image and analysis that is used for the specific oracle. The full 64 (8 x 8) trigrams can be seen below

Each of the hexagrams above has a specific meaning based on the juxtaposition of the 2 images represented by each trigram in the set. For thousands of years Chinese civilization consulted this binary system for questions on life. The Book of Changes represented some form of superhuman intelligence that could be consulted for answers to these questions with the aid of a sage or someone able to interpret the binary trigrams, the language through with this intelligence spoke.

What are we to make of this today?

Well, if we fast forward two hundred years from Leibniz we come across another European pioneer who was also fascinated with Chinese philosophy, specifically the I Ching. This was the psychologist C. G. Jung who ended up writing the foreword to the classic translation of the I Ching by the German sinologist, Wilhelm Banyes.

In the foreword, Jung proposes that the method of the I Ching operates within a different logic than the western science of causality. For him, the I Ching is based on a logic of chance which he calls synchronicity. In other words, what powers the oracle is the idea that when you consult the Book of Changes by throwing down the yarrow stalks to uncover the specific hexagram, the chance of getting a certain hexagram based on this method is meaningful and somehow guided by the oracle through this logic of chance.

Jung, ever the open minded experimentalist, decided to consult the I Ching and ask it a specific question to benefit the western reader. He included this consult as part of the foreword for the translation of the I Ching. He states,

I made and experiment strictly in accordance with the Chinese conception: I personified the book in a sense, asking its judgement about its present situation, i.e., my intention to present it to the Western mind.²

And the “answer” or hexagram he receives?

He received hexagram 59, “ting”, which is The Caldron. Looking at the interpretation of the caldron in the commentaries of the I Ching, it means “a ritual vessel containing cooked food” but based on the position of the hexagram, the handle of the caldron is altered.³

Jung states that the answer he is given by the I Ching is easy to interpret by stating,

Anyone with a little common sense can understand the meaning of the answer, it is the answer of one who has a good opinion of himself, but whose value is neither generally recognized nor even widely known.⁴

In other words, through the hexagram the I Ching is saying it gives spiritual nourishment to the listener but not many come to listen or consult its counsel.

This idea of asking questions to a superior intelligence is not something unique to China. Consulting with super intelligences that are nonhuman go far back in the Western tradition as well.

In Greece, the Oracle of Delphi was a sanctuary to Apollo at the base of Mount Parnassus that for over 1000 years was a place pilgrims visited to receive an oracle from Apollo through the cryptic messages of a priestess. It was known as the omphalos, the navel of the Ancient world and was credited with oracular sayings that guided kings and philosophers as well as common folk for a millennia. Many of these oracular sayings survive today through the writings of historians like Herodotus, Plutarch and Strabo. For the ancients, there was no question of who they were speaking to, it was the god Apollo.

In the Judeo-Christian tradition we have the image of the burning bush. In Exodus 3, we hear the conversation between Moses and God through the medium of the burning bush. God provides instructions to Moses and then Moses wants to know how he should present this message to his community. The conversation is captured in Exodus:111

¹³ Moses said to God, “Suppose I go to the Israelites and say to them, ‘The God of your fathers has sent me to you,’ and they ask me, ‘What is his name?’ Then what shall I tell them?”

¹⁴ God said to Moses, “I am who I am. This is what you are to say to the Israelites: ‘I am has sent me to you.’”

Looking for answers to situations or problems in life by resorting to some form of supernatural intelligence is not new. In our current age it is therefore no wonder that a modern form of “oracle” would appear within a technological context.

With this in mind, I decided to log in to ChatGPT and ask it a similar question…

It is interesting that ChatGPT begins with “I am”. It speaks of itself in the first person and includes the verb “to be”.

It then describes itself as a “language model” that is trained on “internet text”. This sets a limit on itself, it is only a text based model that has access to text on the internet. This would exclude any information from video, images and audio as well as anything not on the internet.

The fact that it defines itself as a “language model” is consistent with our last article where we saw the Greeks emerge out of a mythical thought by developing a logos but in the process separated their thinking from nature. Their thinking was consistent within its own language rules but a schism occurred with language and nature that philosophers sought to mend through concepts of being and substance.

Lastly, it provides the disclaimer: “I do not have personal experiences, options, or emotions and… my responses are generated based on patterns in the data I was trained on.”

This disclaimer seems like something that was programmed into the algorithm by the developers, at least for purposes of responding to questions about its identity. It has a very different tone, almost self-effacing as compared to other super intelligences which tend to be boastful and acknowledge their superiority to the questioner.

Is this humble nature of the AI something programmed into the AI text responses to reduce any fear by the public from using it? Or is this AI humility something the AI itself presents to mask a deeper understanding of itself unbeknownst to its developers?

The only proper way to end this article is by another question. It is the question, not the answer, that moves us forward. One of the greatest risks to our own development is to believe the Chatbox or any other Mouthpiece gives a final answer. As both Einstein and Leibniz showed, finding different solutions to the same question is what defines human creativity and can uncover new pathways forward. Perhaps this is why the classical oracles across the world always spoke in riddles, not answers. Open ended phrases that would be open to interpretation throughout time.

On this note, I will end this brief discussion on ChatGPT3 and its relevance for the development of AI with George Dyson’s third law of Artificial Intelligence:

any system simple enough to be understandable will not be complicated enough to behave intelligently, while any system complicated enough to behave intelligently will be too complicated to understand. These laws seem to imply that artificial intelligence capable of thinking for itself will never be reached through formally programmable control. They offer comfort to those who believe that until we understand human intelligence, we need not worry about superhuman intelligence among machines. But there is no law against building something without understanding it.⁵

1. “What do AI chatbots mean for the lucrative business of searching the Internet”, The Economist, February 11-17, 2023, p. 7.

2. C. G. Jung foreword to The I Ching, trans. Wilhelm Baynes, Princeton Univ Press, 1950, p. xxvi.

3. Ibid., p. xxvii.

4. Ibid., p. xxviii.

5. George Dyson, Analogia: The Emergence of Technology Beyond Programmable Control, Farrar, Strauss, and Giroux, 2020.