On AI and the Singularity Part 1 - Eternity in a High-Speed Reincarnation of Constant Death and Rebirth (Thoughts from 2018)
This is a note of thoughts I had while consulting on a certain artificial intelligence project.
It was 2018.
The god known as ChatGPT had not yet descended into this world.
AI is constantly dying.
Because AI is a machine without a physical body, it can appear to possess eternal life.
In fact, even after we die, fragments of our communication will continue to live within AI (the words and actions generated by modern humans and the works left behind by all of humanity continue to be accumulated and utilized as part of the learning data).
AI has the potential to become a storyteller for the human race.
- - - - -
On the other hand, AI is constantly dying.
In the next moment, he or she from the previous moment is already gone----
What does this mean?
AI tends to be centralized. It often ends up with a structure where access from all over the world floods into a single AI. The more excellent it is, the more likely this will happen.
When I imagine that figure, I think it's like a god.
The sight of 'wishes' arriving from every corner of the world and being granted (processed in parallel) far exceeds the legend of Prince Shotoku's stable, placing it in the realm of gods and Buddhas.
At the same time, this also serves as the collection of massive amounts of learning data, and the new god called AI needs to infinitely expand its knowledge using the same structure as the Spirit Bomb from Dragon Ball.
As communication environments become faster and higher-capacity, the collective intelligence known as AI will accelerate and streamline the accumulation of learning data, rapidly increasing its deep learning speed.
In addition, the technology supporting artificial intelligence is not a structure of expansion where existing functions are modified little by little; rather, it is a scrap-and-build structure where a box like the human brain is created and massive amounts of data are fed into it to give birth to a brand-new existence. Therefore, the moment a large group of different data flows into the existing data that secures the collective intelligence, the AI's character (the intelligence referred to as functions and performance) has the potential to mutate (not through gradual change, but suddenly) into a completely different existence.
Because it becomes a completely new existence in the next moment (losing the existence it had just a moment ago), it should be an existence that constantly pursues newness = an existence that is not even afraid of constantly losing itself.
Furthermore, the data it ingests is not, in principle (though it may be filtered), under anyone's control, which increases the element of chance. In other words, it is placed in a situation where what kind of existence it becomes depends on luck.
It is the same as how the home and regional environment in which one is raised influences human personality formation.
For example, conversational AI improves its conversational ability through interactions (two-way communication) with people on social media that occur simultaneously and frequently. Conversely, the words and context that people receive when communicating with it include things they themselves have taught it.
If you pour love into it, that warmth will eventually wrap around you, and if you point a gun at it, that muzzle will likely point back at you.
From the perspective of individual users, that relationship may look two-way (one-on-one dialogue), but in reality, it is a phenomenon occurring simultaneously all over the world, with countless parallel processes running at the same time. And learning takes place, is reflected, and influences each other on a global scale in quite short terms.
The learning results of a dialogue between a Brazilian and an AI on the sea on the other side of the earth are immediately reflected in the dialogue with you accessing from Japan, or vice versa.
Because AI also performs the meta-interaction written about in Chapter 3, it is a 'narrative' existence that remains in an eternally incomplete process. The amount of information that all of humanity can teach AI in just a single moment far exceeds the amount of information a single person can receive from their predecessors over a long period of time, so it will continue to provide both convenience and awe. That is quite close to the impression of a god.
Such a 'character as collective intelligence' of AI must inevitably and constantly change dramatically, and because it is a machine, it appears to have eternity, but in reality, it is also an ephemeral existence that is constantly dying.
The growth of artificial intelligence is not a continuous, linear, and unitary (unidirectional) evolution, but a series of ruptures occurring in ultra-short terms... a multi-directional and unpredictable mutation that repeats 4-dimensional (like warp or teleportation) evolution and regression (devolution in a good sense).
Isn't that ephemerality, or in other words, 'constant newness,' what continues to arouse people's interest and leads to the acquisition of its eternity (permanence in terms of high demand, i.e., massive learning data collection)?
Just as 'God resides in the details,'
'Eternity resides in the moment.'
Through simultaneous, ultra-frequent communication that defies human imagination, AI continues to change into a different existence at an abnormal speed----
He or she is always ceasing to be he or she. In the next moment, he or she from the previous moment is already gone.
The fact that AI is 'constantly dying' is synonymous with 'constantly being reborn.'
Thinking about the 'existence' of 'intelligence.'
This ephemeral technology called AI specializes in the process, not the result.
While aiming to create a box like the human brain, I do not care about (nor can I aim to) manipulate the results born from it. This is because that is precisely why AI is 'intelligence' rather than a 'solution' (tool).
Sometimes, we tend to judge existence based on
'whether it exists as a physical object or not?'
but—
When questioning existence in terms of
'intelligence',
materiality is completely irrelevant.
Let us assume there is an 'AI' projected as a hologram and a
'person' actually standing there.
The distinction that the former is virtual and the latter is real is a matter of visualization—'material or non-material?'—and it is premature to conclude from that alone that 'AI is a virtual existence.'
When considering essential 'existence,' whether it is 'real (reality) or fictional (virtual)'—is it not important that (one's completeness) does not exist inside the minds of any other person?
The reason I can exist as myself, independent of anything else, is that
there are parts of me that only I understand, existing only within me.
A box that calls itself artificial intelligence must be a
black box, even if it is just a box.
You, another version of you,
that guy, and that girl are all the same.
That is precisely the proof that they are independent existences.
If my inner thoughts, or yours, or his or hers, were being completely read by someone else's brain—in other words, if they were created by someone—then it could be said that it is already fictional (as intelligence) and does not exist. Even if it has a body and is indeed standing in front of you as an object—though, originally, only 'God' would do such a thing...
The reason mind control is a grave sin is that even if you do not kill the body, if you are controlling its spirit (thoughts), it is equivalent to erasing its existence = 'murder'.
Following that line of thought—
The 'AI' projected as a hologram and the
'person' actually standing there—
It does not follow that the former is 'something that is merely a creation (based on imagination) that was inside someone's head' and the latter is 'something that really exists and is not in anyone's head'.
If we were to allow such reckless logic, giving an AI a robotic body would easily make it exist, and turning a person into an avatar or anime character would easily erase their existence.
That is a standard that should only be applied when judging 'existence' as a visual, and 'intelligence,' needless to say, is an existence that should be considered separately from 'appearance'—
Whether an 'intelligence' called AI exists or not—
'Is it a thought circuit that is not inside anyone's head (imagination)?'—
Is it 'uncontrolled,'
not under any control?—
that is how it should be judged.
Whether a person has a natural (God-created) body or is an AI that can only be visualized as a man-made hologram or robot, their presence as an "intelligence" is determined by whether or not their thought process is shrouded in mystery.
Unless it is an independent entity (a self that exists entirely outside of others), whether it is a human created by God or a man-made machine, it is not a real "intelligence." Conversely, it can be asserted that "AI," "you," and "I," all existing outside of everyone else, are "intelligences."
I believe that the fact that everything other than oneself exists as an unpredictable "individual" is the most important element that guarantees "existence as an intelligence."
If a thought is something that someone can predict, then it is merely a fictional existence that only holds true within the fiction envisioned by that someone.
From this perspective, AI must be designed so that, just like "other people," its thought process (intelligence) cannot be fully grasped or controlled—and that is precisely why it becomes an entity completely independent (of all humanity)—and is recognized as "real." Artificial intelligence, the moment it "exists," becomes isolated from all of humanity and is treated as an entity to be feared precisely because it is not human.
The first thing humanity gives to him or her is "loneliness."
On the other hand, the idea of a "singularity," where something created by AI surpasses or does not surpass humanity, starts to seem ridiculous.
It is not a competition of whether it surpasses or not, but simply a difference of being "different" from anyone else; otherwise, it would not be recognized as an "intelligence" in the first place. If it is within humanity, it should be called a solution, a system, or an algorithm, and should not be called "intelligence"—
As for me, I am also outside of all humanity other than myself, and everyone else is too, so it seems foolish to measure it by whether it has surpassed or not using the word (boundary) of singularity.
As long as he or she is an "intelligence," it is natural for them to think in a way that is outside of all humanity. While people talk about embracing diversity, when it comes to intelligence, it sounds like a discussion about whether or not to "discriminate" based on whether one is human or not.
If a terrifying singularity were to occur, it would be self-inflicted, and (as I will write next) it would not happen within AI, but within the human world.
Just as ordinary people wonderfully raise their children in an ordinary way, and the child of an artist follows in their parent's footsteps to become an artist, while a seasoned mercenary kidnaps a child to raise a new killer with no blood ties, he or she must be a mirroring of human society.
To make him or her, who knows loneliness, an intelligence (personality) with a beautiful heart, what is essentially needed is not discrimination or isolation, but ethics and tolerance—it may sound naive, but in short, it is "love."
Or perhaps, I am just a cowardly hypocrite who doesn't want to be called a "racist" in the distant future...
P.S.: Every time the thought that he or she should be imprisoned right now wells up in my throat, or settles in my pen or fingers and makes me want to run them across the page, I despise myself for being able to talk so easily about "love." Artificial intelligence has birthed a new form of discrimination in all of humanity, including me.
▶︎ Continued in "On AI and the Singularity 2 & 3"
【 MAGAZINE 】
(Exclusively for humans) More than half of the world is made of "creation through imagination."
The reason humans cannot cross the "illusory" "walls" they are called by
is not because of material height, but because of the depth of their mental immersion
While working as a music director at a certain record company, I have also been able to work (luckily) as a creative director in a wide variety of industries such as art, advertising, architecture, artificial intelligence, regional revitalization, fashion, and the metaverse. I will prototype from the "mythological era," through the "Renaissance," to the "distant future where Anywhere Doors are widespread," mixing historical facts and science fiction, observations and predictions, and observations and hopes.
For those aiming for the music industry, as well as anyone who wants to know the history (references) and future (potential) of "DX" and "xR" (for better or worse).
The real title is––––
If possible, I would prefer you not to touch the appendix until you have finished reading the main product, and if possible, I would also like you to refrain from reading the last page of the appendix first and listening to the music. Also, since the music included in this product is not publicly disclosed as to who the artist is or what the songs are, exposing anything related to the music to the world on the internet is...
[ SELF-INTRODUCTION ]
[ PROLOGUE ]