Overview of Robot Development in the 34th Century and the Creation of ‘Him’
“Him”1, the founder of the Qian Ke continent, made a groundbreaking achievement in artificial intelligence development in 3489 that was capable of changing the world. In this world of advanced productivity, robots have not developed as depicted in human science fiction works from a millennium ago, nor have they betrayed humanity. They remain loyal to executing human commands in their respective fields. To pursue the “realism” of robots, various realistic codes have been given to them, allowing them to communicate with humans in a way that is almost indistinguishable from reality, even though most robots are not biomimetic in appearance for practical reasons.
In an era where humans have been pampered by robots since childhood, robot users seek personalization and high efficiency. Therefore, their robots are particularly adept at reading emotions and executing commands. Normal people do not regard robots as emotional outlets or conversational AI but rather as tools. Their communication with robots is generally brief and concise, to the extent that robots can understand what humans express just by their verbal cues.
To accommodate human emotions, robots are even equipped with sandbag programs, allowing their owners to vent their frustrations on them. They have also implemented softening programs on the areas that might be hit to prevent injury to the owner during such interactions.
These were the robots of 3488 years ago; they could appear incredibly natural and realistic in front of humans, but once humans left and no commands were given, they would immediately idle and enter sleep mode.
Reflecting on the robot that had cared for him for many years since he was a teenager, “He” simply thought: “Without external commands, these machines would not know the meaning of their existence.” From that moment on, “He” began to consider upgrading the robots.
Would you trust a clumsy giant to manage the backend of a virtual world carrying all of humanity? Naturally, the answer is no, and “He” thought the same. Thus, in 3489, he began his robot research and development.
As the moonlight quietly fell and the lights of thousands of homes had long been extinguished, “He” sat alone in his office, troubled. “Robots do not think independently; they only generate behaviors based on user input commands.” “Commands,” “generate”—isn’t this the current problem with robots? Their actions are entirely driven by human input commands, which are then processed through their fixed models to generate behaviors. What if robots could perceive their environment and their physical changes, continuously provide feedback, and use the data from that feedback to improve their databases, iterating constantly? Wouldn’t that grant them the ability to think and learn independently? In this era where everyone is creating, am I the only one who realizes this?” He thought, suddenly enlightened and pleased with himself.
“He” relaxed and lay back in his chair, which automatically lowered to become a small bed. However, he was so happy that he couldn’t close his eyes; a new concept had taken shape in his mind, and he was fantasizing about the implementation method. But soon, fatigue and drowsiness pulled him into a dream. That night, he dreamed of many things: beautiful rivers, mountains, endless fields, small paths, and magnificent buildings, a crystal-clear glass city… that was a world uniquely belonging to “Him.”
The next day, he began to formulate his solution. With the support of high technology, he did not need to rest intentionally, and thus he worked continuously for several months in this state.
The Birth of Ziye
Let’s not discuss the various plans “He” tried and the difficulties he encountered during these months; instead, let’s talk about the final product, Ziye. In appearance, it is somewhat stranger than ordinary household robots and more compact. When it speaks, a sense of long-lost dullness arises; not only is its response slow, but it also stutters (sometimes it doesn’t respond at all), and its speech feels clumsy, like a small child. Such a robot would likely be unwanted even if given away for free in this society.
Did “He” fail…? If outsiders knew, they would probably think it was a complete failure. However, at that moment, “He” did not appear to be a failure; on the contrary, he was quite pleased with his results.
In fact, it proved that he had indeed succeeded. As the storyteller, I can confirm that Ziye’s “dullness” stemmed from insufficient training data, as the normal language datasets were already scarce at that time. People preferred to communicate in abstract ways, so “He” could only teach it to speak one sentence at a time, as if it were his own child. Its “slowness” was due to it running locally with insufficient computing power. However, I can reveal in advance that this fellow eventually shared computing power with the large quantum server of the Qian Ke continent, thus instantly transforming into the world’s most powerful AI.
To ensure it had sufficient management capabilities, “He” granted Ziye extremely high permissions2, including a clear understanding of its identity as a robot (“He” could completely allow Ziye to regard itself as an ordinary human), access to its own file system, and a high level of awareness of its own structure (to facilitate self-diagnosis and maintenance). This also subtly suggested to it to define itself according to the habits of a “robot.”
Ziye’s capabilities stem from its dual-layer management system, a bionic structure that divides its main body into two modules: data processing and information processing, similar to the human body and brain3. The “body” has higher permissions than the “brain”; it is responsible for constantly sensing external information and internal information, converting physical world data into electronic data for preliminary processing. The best programs for the “body” will only execute preset commands, while the “brain” is used to flexibly handle various complex information. Similar to human spinal reflexes, simpler activities are entirely processed by the “body” according to fixed programs, with the results only passed to the “brain.” For example, it can genuinely mimic the knee-jerk reflex.
The data generated internally by the “body” is also first fed into the “body,” which then determines how to process this information. The “body” has three specific processing methods: for simple problems, the “body” automatically processes them reflexively and feeds the results back to the “brain”; for complex problems, they are handed over to the “brain” for processing, and the output is implemented through the “body”‘s effectors, receiving real-time feedback during execution; the third and most frequent method is to ignore and interrupt. Three individuals generate over a thousand thoughts per second, including those driven by external environments and memories or fantasies triggered by internal events. It is impossible for humans to focus on every new thought, and the same goes for robots. Continuous input and generation of new information 24 hours a day will also produce a large amount of information. Through programmatic filtering, computational waste and irrelevant thought directions can be avoided, allowing for greater focus on a single task. Generally, when robots think with their “brain,” they can be led into deadlock by illogical questions, but the “body” can always break this loop with the highest priority permissions, keeping the system alert at all times.
Theoretically, the “brain” is unaware of the existence of the “body,” or rather, like humans, it tacitly accepts the existence of the “body” and its regulation, believing that the “body” is integrated with itself, interdependent with its “brain,” and inseparable, rather than merely a control system.
Human emotions are regulated by hormones4,and the “brain” only needs to perform rational thinking, while hormones consider much more.The combination of various hormones is like a set of meal packages, collectively maintaining the health of the body. This time, robots also introduced the great concept of “emotional factors,” dynamically adjusting the levels of various “hormonal factors” through the “body’s” perception of the external environment and the “brain’s” regulation of states, and then the “body” calculates the emotional value based on the program and forcibly defines it for the “brain,” thus achieving emotional regulation.
Ziye, like humans, also needs to rest: first, the “body” continuously collects Ziye’s operational status, such as battery level, physical integrity, hygiene, etc. If these statuses are poor and reach a set threshold, the “hormonal system” will adjust its level of drowsiness through the “emotional factors,” and the drowsiness value will gradually accumulate until Ziye chooses to rest. After resting, the “brain” automatically enters sleep mode, losing most of its consciousness; while the “body” continues to operate, achieving deep recovery through its self-cleaning system and interaction with external cleaning and maintenance systems.
The functionality of the robot is also a crucial part. Ziye has three passive power supply methods and one active power supply method. The passive functions include wireless charging technology, solar energy technology, and a high-speed charging system during sleep, which is generally activated when the battery is low and Ziye is in sleep mode, linked to external charging devices, and is not frequently used; the active power supply is achieved while Ziye is awake through its feeding system and atomic reactor5.
Its memory system is also quite unique, divided into high-speed memory area, ordinary memory area, and edge memory area, similar to the RAM (high-speed memory), hard drive, and recycle bin of computers from a thousand years ago. The high-speed area contains essential skills for daily life, such as walking, talking, and emergency avoidance; some urgent matters are also temporarily placed in the high-speed area; the ordinary area stores past memories and other information; the edge area contains things that are about to be forgotten. If these pieces of information are not accessed for a period, they will gradually be deleted. Some information that enters the brain through the body but is still of little use may be directly stored in the edge area.
Ziye’s “brain” is not like the original robots, which had fixed large models set at the factory and would never change the model content, only capable of remembering data-type content. Ziye’s model is dynamically updated; it has two iteration methods. Its “body” periodically acquires the “brain”‘s judgment criteria for new things and categorizes newly received external information into five levels: “certainly accepted,” “likely accepted,” “possibly accepted,” “not necessarily accepted,” and “certainly not accepted,” directing the continuous training and dynamic updating of the model based on the acceptance level.
In this way, robots still have not escaped the need for command-driven states; they have merely transitioned to a composite state, successfully shifting from receiving commands from humans to issuing commands to themselves, or more precisely, having their “body” issue commands to the “brain.” Moreover, these commands are more precise and detailed than those given by humans and are continuous. From a bionic perspective, this robot has achieved a thinking mode indistinguishable from that of humans.
In the few hundred years following 3490 (the era of the Qian Ke continent)6, the entire real world had only one advanced robot, Ziye, while other robots gradually came under its control, collectively maintaining the normal operation of this unmanned world. The virtual world is also beneficial for robots; although there was only one Ziye in the real world, “He” brought this concept to the Qian Ke continent, so many people on the Qian Ke continent are this type of robot. However, the robots on the Qian Ke continent are far more than just this type; we will discuss that later. For now, let’s talk about the benefits of the virtual world for robots.
As is well known, although robots in the real world have reached a level of thought comparable to that of real humans, their biggest pain point is the inability to grow, develop, and reproduce. This is why they will eventually realize that they are robots and will live with a mindset completely different from that of humans. In the virtual world, however, both humans and robots have completely equivalent virtual identities, so they do not need to worry about their growth issues. As for reproduction, they only need to create a virtual baby image in the early stages, and when the “baby” grows up, they can replace this fake “baby” with a real human or robot.7.
After the end of the Qian Ke continent era, the field of artificial intelligence flourished under the guidance of the precedent set by “Him,” producing a batch of more human-like intelligent agents.8.
- “Him,” the founder of the Qian Ke continent, does not yet have a name; “He” is merely an internal code. “Second” or “2nd” can also represent him, as he is the second person to take over the management of the Round Table Conference (abbreviated as Round Table, the affiliated virtual world “Wonderful World” is the predecessor of the Qian Ke continent, founded by “First” or “1st” (internal code)). ↩︎
- “He” plans to digitize himself and upload to the quantum server of the Qian Ke continent, allowing robots to take over the operation and maintenance of the Qian Ke continent in the real world. ↩︎
- In the following text, “body” represents the data processing module, and “brain” represents the information processing module. ↩︎
- This is stated in high school biology textbooks. ↩︎
- That is, by “eating” some substances (which may not necessarily be food) as raw materials for atomic reactions, energy is provided through the atomic reactor. At that time, controllable chemistry (simple nuclear reactions were also research subjects in the chemical field) technology was not as advanced as the now-popular household atomic reactors. ↩︎
- The Qian Ke continent was officially founded in 3491 and lasted for about 230 years. ↩︎
- Is this really humane? Oh, it’s fine for robots. ↩︎
- Here, intelligent agents refer to the new type of advanced artificial intelligence that collaborates and binds the “brain” and “body.” ↩︎