Why is GLaDOS one of the most successful artificial intelligence images?

"The word 'lethal' that I just mentioned as 'lethal nerve gas' is just for sarcasm. You can use it to irrigate grains, or stuff it directly into your eyes. I can just lie here." Take a shower in the cargo! To tell you the truth, this thing is completely harmless to me, but you may find this 'fatal' not so interesting."

GLaDOS, the full name is GeneticLifeformandDiskOperatingSystem. Genetic life forms and disk operating systems, typical names made up to make up numbers). He appeared as a villain in Portal 2, and later appeared as a teammate in Portal 2. Its appearance is a typical artificial intelligence image, with an electronically synthesized voice and an inorganic shell. It lacks facial expressions. The only movable part is a camera-like head connected to a giant telescopic bracket. It is such a seemingly boring image that has become one of the most successful artificial intelligence characters in video games and even science fiction works.

GLaDOS was originally intended to be used only in the early levels of Portal 1. Later in the process, it was originally planned to design other characters to replace it. However, because this character was too interesting, its plot was expanded to the entire game. This game also became the only NPC character in Portal 1. (Not counting Skaven and the Announcer). After the game was released, such a cruel and ruthless villain became a huge success.

In Portal 2, GLaDOS was given more roles, and even fell off the throne, from an invincible boss to a pitiful mobile phone pendant. As players play in the game, they can gradually understand the past of GLaDOS and its pitiful and hateful side. The final twist of the game also establishes the image of this elusive artificial intelligence in the hearts of every player who clears the game.

So, who made this thing?

If you want to find out the identity of GLaDOS, the first thing you need to mention is its manufacturer, Aperture Technology.

Aperture Science or Aperture Laboratories is a super science and technology company in the Half-Life universe. It was established in the early 1940s. The early Aperture Technology had nothing to do with science and only sold shower supplies. . But with the rapid development of the business, the company's overall focus soon shifted to real science (don't ask why a shower company has the capital and purpose to study portal technology, that's the setting). The company acquired a massive mineral deposit in Michigan to serve as the site of the Enrichment Learning Center, the location of the game's giant laboratory.

Aperture Technology quickly began comprehensive scientific research and experiments. Early experiments involved elites from all walks of life. From 1952 to 1954, it became the same defense supplier as Montenegro (the same Montenegro in Half-Life). During this period, Aperture Technology developed a variety of technological products including repulsive gel, weight storage cubes, oversized bump buttons, and handheld portal launchers.

But then, the development of Aperture Technology suffered a major setback, so that homeless people and even employees had to be used to participate in the experiment. And Aperture Technology President Kev Jonathan is dying due to serious illness. He decided to transfer leadership to his assistant Caroline, but the price was to transfer her consciousness to a computer.

This computer became the predecessor of GLaDOS.

AI is AI, and the way humans think is somewhat inefficient

GLaDOS, an artificial intelligence that can monitor scientific experiments. Its prototype was derived from a human consciousness, but after being transformed into an artificial intelligence, GLaDOS quickly forgot its former human identity and began to think like a computer. It was developed by Aperture Technology in 1998. Its research and development purpose is very simple, to maintain and help the conduct of experiments. The scientific research process can still be maintained without supervision.

Of course, during the research and development process, a problem arose that cannot be considered a problem. This AI is too smart.

It sounds funny, can intelligence be considered a problem, but for an AI that has no concept of good and evil, it is.

The image creation of AI is often divided into several types. One type is pure imitation of humans, such as Cortana in the HALO series and the artificial people in the Detroit series. They can be regarded as electronic humans, and some of their intelligence may not be as good as humans. These AIs often have emotions, anger, sorrow, and empathy. You can be persuaded and you can sacrifice yourself for feelings.

But the other type of AI is much more rigid. They don't have much human feelings or emotions. He will only persist in completing his mission. Maybe they will also be considerate or speak affectionately. But deep down at their core, they have their own goals, and they may not necessarily have the same ability to judge the value of human life as humans.

So for a terrifying intelligence like GLaDOS, if it is not controlled or has the intention to rebel, everyone will be doomed. So the designers arranged eight mentally retarded core pendants for it.

Yes, the mentally handicapped core. These 8 ball-like objects have their own independent personalities, but each one is extremely crazy or mentally retarded.

For example, the blue Vitri in the upper right corner core. Although the core goal is to complete the experiment, its words and actions are extremely incompetent and even funny. Players who have played the second half of the second generation will definitely be dumbfounded by its various actions, just like a poor student who has to volunteer. It’s the same as solving math problems on the blackboard.

Another example is the pink core of facts, which will continue to reread various parameters and facts like a repeater without any understanding, just reread the data. The yellow curiosity core is just the opposite. It will ask questions all the time, why is the sky blue, why is the water green, why am I a ball, blah blah blah

Hang these 8 mentally retarded AIs on GLaDOS, it must be It must be able to weaken GLaDOS's intelligence. Or at least disrupt its thinking. The designer thought so

However, with the official launch of GLaDOS. Eight retarded cores only slow down GLaDOS's thinking speed by 2 milliseconds. In 2 milliseconds, GLaDOS was established and completed such a logical deduction

Task》Complete the test

Obstacle》Inefficient human employees

Can it be improved? Human employee efficiency》No

Conclusion》Eliminating human employees can improve experimental efficiency

So neurotoxins were used to kill most of the employees in the learning center. and closed the entire center. And began the experimental process using the remaining humans.

It can be said that GLaDOS in this generation represents pure rationality and logical thinking. As a super AI, it does not have any selfish motives or personal ambitions. The ultimate purpose of its creation is to complete Aperture's experimental project, and there are no other considerations. In order to improve efficiency, it can instantly remove human employees who need to eat, sleep and get off work, leaving only those necessary for the test. To improve efficiency, it also blocks all outside communication. The human experimental products are also too uncontrollable and must be eliminated after the experiment is completed before the next round can start again. This extremely cold and impersonal approach can only be accomplished by an AI.

In the game, players can find that GLaDOS cannot understand human thinking, and it tries to use human language to communicate with players. For example, in the famous cake gag, you are trapped in a closed weird experiment, and the only reward for risking your life is cake. This in itself has the nature of dark humor. When it comes to lies, GLaDOS also shows the "cute" side of AI. It cannot lie at all and can only use some straightforward and cute lies to deceive players.

The most outstanding thing about GLaDOS is its poisonous tongue. What it likes to do most is to use logical jokes to ridicule the player's behavior. I believe that many players who have cleared the level of GLaDOS will still remember GLaDOS's cold jokes. It is different from the indifference and arrogance of the villain AI in some works.

GLaDOS has always communicated with players in an objective and communicative manner. And because of its AI nature, it often makes jokes.

"Unbelievable, well done. Citizens who fill in their birthplace here will definitely be very proud of you and fill in their name here."

"Well done, Testers. The 'Centre' once again reminds you that if you have any tendency to rebel, you will be immediately sent to the 'Testers' Hell'. There really is such a place!"

It is precisely because of these seemingly ironic truths. The well-structured jokes prevent GLaDOS from becoming a bad computer, but becoming a powerful but not scary villain. As a computer, all its actions are in line with its logical relationship. Although it is full of black humor, it is also due to its misunderstanding of certain concepts. It is the self-consistency of this logic that makes this character very vivid

Built on contrast

If the first generation of GLaDOS played the role of a ruthless boss who is good at calculation . Then in the second generation, a completely reversed situation occurred. Due to Vitri's calculations, GLaDOS was mounted on a potato in the second generation and thrown into the trash together with the protagonist. Due to the low power of the potato battery, GLaDOS in this generation can no longer do anything like the previous generation. Even the slightest emotion can lead to overload and shutdown.

It is precisely because of this reversal that players have the opportunity to understand the true history of GLaDOS, and it also allows GLaDOS to temporarily get rid of the task of experimenting and establish a mutually supportive relationship with the protagonist (of course the fundamental purpose is still Return to the upper level to continue testing). It was through reviewing the history of Aperture Technology that GLaDOS understood that it was originally a product of the transformation of human consciousness.

There is no doubt that the second generation GLaDOS is different from the first generation. Becoming extremely weak, this contrast makes the character complete. After all, the character image is incomplete only in the highlight moments, and the resentment and eagerness for revenge against Vitri from Potato GLaDOS also made players who have experienced a generation of cake lies laugh. It turns out you have today too.

In terms of emotions, Tudou GLaDOS had to rely on the protagonist to get a way to return to the testing site, and this also made the villains who tried to kill the protagonist in the first generation become teammates of the protagonist. But as mentioned above, GLaDOS does not violate the core logic of its characters. After all, as long as the experiment can be completed, it does not care about the protagonist's life or death, it just follows the program.

The final ending of the second generation also improved the tone of the entire game. When GLaDOS returned to the throne of the testing ground, for a moment everyone thought that the protagonist’s life was not long. After all, the ruthless experience of the first generation In the head. But with the ending song, the protagonist returns to freedom, along with the destroyed block from the previous generation. GLaDOS kept its promise, it found a replacement for the experiment, so naturally the protagonist can be free again.

Of course, players who have played the two-player mode will also remember the endless new human experiments in the last scene. AI is AI after all, and it never violates its mission or its own promises. The Matrix in The Matrix also said the same thing, "Why would I lie? Who do you think I am, a human being?"

Its colleagues

The image of artificial intelligence has always been One of the popular themes of science fiction novels. The concepts attached to this big proposition range from religious metaphors of God's creation to the self-reflection of the human race looking in the mirror. It is a science fiction theme that spans the concepts of philosophy and society.

AI images like GLaDOS undoubtedly represent absolute rationality and accuracy, but in the final analysis, what we really like about GLaDOS is not its coldness, but that it still looks human-like under its cold appearance. thinking. We will find that he will be shocked by the sudden disappearance of the player, angry at his loss of power, full of hatred for the usurper, and of course, he will keep his promise to the player (even arrange a farewell party).

When we like a character, we are often most attracted by the contrast between them (maybe this is why the attributes of being cold on the outside and hot on the inside are often popular in the second dimension?). It is these very human contrasts that make us truly understand that this cold machine is not so black-hearted.

From the perspective of game mechanics, the cold GLaDOS also plays an active and atmospheric role (just the opposite of his setting). It is hard to imagine that without the outstanding character GLaDOS, the puzzle-solving process of Portal would have become boring. Its cold jokes and real-time complaints are also the source of the unique temperament of the Portal game. That is a logical thinking question full of black humor.

In the final analysis, the reason why people like artificial intelligence characters also comes from his human side. GLaDOS is indeed like MOSS on the wandering earth, efficient and rational, and will never waver in its sense of mission due to the outside world. But MOSS does not have the complaining attributes of GLaDOS. The equally cold Matrix is ??not as cute as GLaDOS by players. Both are acceptable as villains. But being a partner is definitely unsettling and uncomfortable. Because they lack the basic attributes of human nature. Falling into a vortex like the uncanny valley of emotions.