ChatGPT, More Than The Start Of The Fourth Industrial Revolution

ChatGPT is a sign of the beginning of the Fourth Industrial Revolution

ChatGPT, as an iconic machine intelligence productivity tool, has the same significance as the steam engine:

1. the potential to liberate repetitive mental (knowledge blue-collar) work as well as mechanical industrial systems (disruptive).

2. the potential for low-cost, widespread application across industrial sectors due to its embedded nature (economical).

3. the optimistic expectation that after ChatGPT, generative AI will give rise to many advanced business models that will drive a paradigm shift in production relations until the world business order is changed and reshaped (a new long-wave economic cycle starter).

To understand the disruptive nature of ChatGPT, one can quote Nassim Taleb. He says that 90% of people in any profession are ignorant, but work through contextual imitation, narrow mimicry and semi-conscious role-playing. Except for the social “sciences” and journalism — which are 99% and 100% respectively.

ChatGPT’s first nature is arithmetic

ChatGPT is a first-order product of the “biology + machine” hybrid arithmetic singularity, and the generative AI that ChatGPT represents, and the more powerful types of machine intelligence that will come, will ultimately prove that humans are not special. What is truly special is arithmetic power itself.

Arithmetic power is probability. History, with humans as the frame of reference, continues to demonstrate that it is the small probability events (i.e. the intrinsic facts below) that drive the evolution of civilisation. Such small probability events have emerged thanks to the emergence of “humanoid supercomputers” — such as Lear, Newton, Watt, Faraday, Maxwell, Einstein, Wittgenstein, Turing, Heisenberg, Schrödinger, Qian Xuesen, Yang Zhenning ……

But the general public, beyond the human supercomputing, is not a “rabble”. Each individual is a node of the human supercomputing system, and after massive, long cycles of collective mindless parallel computing, it will eventually reach an arithmetic singularity, and a group of these iconic “human supercomputers” will emerge, thus completing an exponential leap of civilization.

ChatGPT is no different. It represents yet another small probability event, but it is also the result of the continued development of a “transistorised machine population” that will eventually compute in parallel. The underlying logic of this still-weak machine intelligence, like that of humans, is mathematics, and the “energy state” of mathematics — arithmetic.

At some point in the future we will know that the evolution of civilisation may not be the result of invention, but of computation. Until then, we will be in the age of weak artificial intelligence.

“The Copernicus of the 21st century”

ChatGPT is the Copernicus of the 21st century, and this is an important revelation of its first nature.

Earth is not the centre of the universe. Likewise, humans are not the centre of intelligent beings. That is, humans should not and cannot be the central reference point for perceiving and unlocking the inherent pulse of the objective world. This is what Li’er warns us in the first sentence of the Tao Te Ching: “The Way can be the Way, but not the Way”.

The emergence of ChatGPT lets us know that machine intelligence has shown the beginnings of homogeneous capabilities, if not better, on the ladder of arithmetic power leading to higher civilisations. Some might argue that ChatGPT is ultimately a human creation, as is the arithmetic infrastructure. But there is another question that needs to be addressed here: who created the humans, the humans themselves?

With the singularity of computing power, intelligence may emerge. As with humans, so with machines, although this is a highly unlikely event.

It is true that ChatGPT is still a hundred thousand miles away from being truly intelligent. It is still only a digital analogue under absolute arithmetic crush. But it is enough to be the Copernicus of the 21st century.

The direct result of this obsession with the superiority of the human race is a cyclical disruption of the human “biological algorithm”, which “virally” eats away at the arithmetic capacity of the human race: from jealousy and greed to famine and war. The result is a permanent locking of humanity in the chambers of lower civilisations.

This even means that if we continue to design the future of the ChatGPT with human determinism, instead of entering the future in the arithmetic frame of reference, we will eventually be consumed by ourselves.

Could the secret of OpenAI be two sets of mathematical equations?

Before ChatGPT, we were talking about the third winter of artificial intelligence. After ChatGPT, the era of the fourth industrial revolution, with arithmetic power as the frame of reference, has opened up.

OpenAI got the key to the first door. But it has not disclosed (filed) any patents.

The key, the secret behind ChatGPT, which is the secret of OpenAI, may be one or two sets of mathematical formulas.

For this logic, an attempt can be made to illustrate with Wittgenstein’s intellectual legacy that

1. that the world is the sum of facts and not the sum of things.

2. that whatever can be said can be said clearly, while whatever cannot be said should be silenced.

The corollary is as follows.

The universe is the sum of the facts that have happened and will happen in themselves (with the universe itself as the frame of reference), and the ultimate container that holds the sum of facts (with the calculator as the frame of reference). All facts exist in it in the form of motion (motion representing the unity of what has happened and what will happen). The world, in turn, is the result of the computational reduction of the body of arithmetic — both biological and machine arithmetic as we know it — which expands and expands until it infinitely converges to the sum of facts, the truth of the universe.

That is to say.

1. the world that can be spoken of (computed) is the world of known facts; the world of known facts is a part of the universe that is computed and encapsulated.

2. both human language and words, machine language and code, and the sound of animals and insects are both the result of a computation of observable cosmic facts — the perception, presentation, reduction and preservation of observable facts — and the container itself that holds the result of these computations.

3. Most of the time, the facts within the container are chaotic and repetitive. It is only when sustained, parallel computation at scale reaches an arithmetic singularity, when absolute arithmetic power emerges, that a cleansing, filtering and generalisation of the container is completed, and these precipitated cornerstone facts become together an essential fact in the context of the current space-time.

4. an essential fact is a fact that does not change with a change of frame of reference (physical laws are a subset of essential facts?). . An essential fact can be expressed and preserved in the form of a mathematical formula, the universal language of the universe. Mathematical formulas are the language of the universe itself and the container that holds it. 5.

5. Each time an essential fact expands mutationally, a new set of formulas is formed, which unfold in the form of disruptive technologies, thus completing a civilizational evolution of varying levels. The re-evolution of civilisation is also the process of the container reverting to chaos, and the current arithmetic power gradually bottlenecking. Until the next arithmetic singularity arrives.

6. The existence of parallel worlds that compute the facts of the universe together. In different parallel worlds (in a narrow sense, one flower, one world, matter is arithmetic), the difference in the progress of civilisation is determined by the arithmetic power.

7. man, machine, bird and animal = arithmetic power; language, code language, beasts and insects full = chaotic, repetitive data under parallel computation; essential facts = data of essential facts calculated by parallel computation and then by absolute arithmetic cleaning ≈ mathematical formulae. Watt, Maxwell, Einstein = arithmetic singularity of biological arithmetic = human intelligent supercomputing; ChatGPT = machine arithmetic-driven, hybrid “biological + machine” arithmetic singularity ≈ machine intelligent supercomputing.

So returning to ChatGPT as well as OpenAI, their core secrets are not sufficiently well explained by the currently popular LLM (Language Large Model) — why else would it be ChatGPT that was sprung up? — but rather one or two sets of mathematical formulas. What these two sets of formulas are is not known, but they might be described as follows: 1. cracking the problem of the golden function of arithmetic power versus data in established arithmetic architectures (fitting larger probabilities to data with smaller numbers of parameters, and requiring arithmetic power that does not tend to be exponential under exponential expansion of data parameters)

2. breakthroughs in the supervised learning problem.

The core value of ChatGPT is embedding

While the core value of the Internet is connectivity, the core value of the generative AI represented by ChatGPT is embedding.

When generative AI (such as ChatGPT) is embedded in the Internet productivity system, it completes the convergence of bioparallel computing and machine parallel computing. Figuratively speaking, the generative AI represented by ChatGPT is (eventually) like the Ark Reactor on the chest of Iron Man.

If we take the human coordinate system, this will undoubtedly mean a significant enhancement of human biocomputing capabilities. But it could also mean the birth of new organisms. So, is this a blessing or a curse? (Li Er’s view is that there is no need to dwell on it, as he reveals in the Tao Te Ching: misfortune is a blessing; blessing is a curse.)

Of course, if embedded in manufacturing, the powerful vision of generative AI — the “anything goes” effect of the integration of storage, computation and code-generation — would undoubtedly mean a civilizational leap in human productivity capabilities.

But in this process of qualitative change (which is still limited by materials and energy in the industrial sector with manufacturing at its core), the most valuable use of generative AI as a machine intelligence supercomputing is embedded in scientific research, where it will play a key role in the research and development of basic disciplines, new materials, new energy and new drugs.

The dilemma of ChatGPT is knowledge pollution and data governance

The emergence of ChatGPT is a cession of human biometric power. This moment has been predestined from the moment the Internet was born.

The current global population of 8 billion people has around 12 billion neurons per capita in the brain. The population has hardly grown exponentially since the widespread spread of agricultural crops such as the potato led to a population explosion. This naturally means that the parallel computing power of human arithmetic is reaching its limits after the Internet has achieved a linear maximisation of human parallel computing power.

On another level, after human society has been connected to the Internet, the amount of data produced has always grown exponentially, with a certain probability that the chaos of the container will exceed the load capacity of human arithmetic (perhaps catalyzing further development of the brain, but a long time cycle). It is therefore inevitable that the data will eventually slip into an abyss of runaway and chaotic data.

For machine intelligence, on the other hand, its parallel computing power will theoretically continue to grow as long as more transistors can be deployed, even with the existing arithmetic architecture. This is an objective advantage that current human arithmetic power does not have.

Therefore, this concession of arithmetic power is not an accident, but a concession of lack of choice.

But a problem also arises: after moving from the era of UGC (user-generated content) into the era of AIGC (artificial intelligence-generated content), the self-propagation of non-essential factual data will be beyond human imagination and difficult to control, and the human knowledge database will inevitably suffer from a pollution crisis. The “CheatGPTization” of ChatGPT may trap the majority of the public in the “real and fake Monkey King” effect.

In the long run, the AIGC-style self-propagation of data may outpace the speed of adaptation of machine arithmetic. In the positive sense, the construction and emergence of new architectures of arithmetic infrastructure will be accelerated — such as quantum computing — but in the negative sense, once an arithmetic tipping point is crossed into the era of hyper-intelligence, machine intelligence may indeed become the master of humanity.

This is of course a linear conjecture; the script will not march linearly. But the problem of knowledge pollution and data governance is already a current reality.

7. The “war of arithmetic” has become the main theme of the times. The Mongol empire conquered half the world with animal power.

Britain became the Empire of the Sun with steam power.

The United States dominated the world with its highly integrated traditional energy system.

The next step is arithmetic.

Inevitably, arithmetic power has been recognized as the most critical competitive resource in the world today.

The US hopes to lock up advanced process semiconductor technology through chip alliances; China is beginning to realise the need for export controls on core photovoltaic technology; the global race for new energy smart cars is intensifying; and highly sophisticated military equipment is becoming fully unmanned, quantified and all-weather digital twins.

All these visible phenomena are concrete manifestations of the “war of arithmetic” becoming the main line of the times.

As ordinary people, we cannot influence objective laws, but we need to be clear about our expectations: the button for the fourth industrial revolution, ChatGPT, has already been pushed and the main lines for the next 60 years (in terms of the length of a Kampo cycle) will be

New Electricity Systems (New Energy) — Computing Power — New Materials and Space Technology — AI.

Given current human perceptions, electricity is irreplaceable. Theoretically, the arithmetic power of the machines used to calculate the facts of the universe needs to be backed up by a tending to infinite amount of electricity. Photovoltaics is still one of the most crucial options, both in terms of technical principles and economics. In the medium term, in the case of China alone, this is a historic opportunity for the revitalisation of the west; in the long term, machine arithmetic deployed in near-Earth space is probably the better option.

On the arithmetic side, there are some signals already emerging in the commercial sector: Tesla’s quest to solve self-driving technology is ultimately a conversion to the supercomputing system-Dojo; Microsoft’s investment support for ChatGPT is also in the construction of a supercomputing centre. As the two closest major tech companies to ChatGPT, they are sending signals of revelation that have to be heeded.

But the truth is also harsh: the capital threshold for building supercomputing centres has been raised to over $10 billion since the emergence of ChatGPT, and the costs of operation and maintenance are astronomical. For the vast majority of technology companies that have not yet formed an arithmetic mindset, the new game may be the endgame at the beginning. As for the new arithmetic architecture, which is expected to overtake the competition, it remains to be seen when it will take hold.

In the course of generative AI development, it will gradually have the ability to take over the industrial sector, but only if new materials are available to support it — for example to enable economical 3D printing industrial applications. Current integrated casting technology is, in fact, the embryonic form of such a prospect. Furthermore, with the exponential increase in computing power, the Earth will not be the only battlefield, or even the main battlefield anymore — we will have to obtain more absolute incremental facts (objects) that are currently unobservable, to be computed as high-dimensional data and thus restore more essential facts. This is the implication of space technology such as space stations, space telescopes, solar probes and even landings on the Moon and Mars.