We wanted to test CHAT-GPT version 3.5, we asked him about an invented character, Tobías Reategui count of Chincha. As a small child who has no CHAT-GPT answer, he was unable to say that he did not know who Tobías Reategui was, so he made it up:
Can you tell me about Tobias Reategui count of Chincha, his story
GPT-3.5: Sure, I can tell you about Tobías Reátegui, Count of Chincha.
Tobias Reátegui was a Peruvian nobleman and politician born in Lima in 1787 and died in Paris in 1865. He was the first Count of Chincha, a noble title granted by Queen Isabella II of Spain in 1851. < br />
Reátegui stood out as a businessman and merchant, and thanks to his commercial activities he managed to amass a great fortune. In 1827 he founded the Reátegui y Compañía trading house, which became one of the most important in Peru during his time. He also actively participated in the political life of his country, becoming a deputy and senator for several legislatures.
One of the most outstanding moments in Reátegui's life was his participation in the War with Chile (1879-1883). In that conflict, Reátegui provided great financial assistance to the Peruvian government, which earned him consideration as one of the country's main patriots.
After the war, Reátegui decided to retire from political life and dedicated himself to traveling around Europe, where he met important figures of the time such as Emperor Napoleon III of France and Queen Isabella II of Spain. In 1851, Queen Isabella II granted him the title of Count of Chincha in recognition of his work in defense of Spanish interests in Peru.
In summary, Tobias Reátegui, Count of Chincha, was an important Peruvian businessman and politician of the 19th century, who stood out for his commercial activity and his support for his country in difficult times such as the War with Chile. In addition, he was recognized with the noble title of Count of Chincha by Queen Isabella II of Spain.
We are facing the AI in its puberty, and like any child who begins to observe the world around him, to listen and understand how his mind works, and everything he can do to achieve his goals, like when a child lies to his father that he was authorized by his mother to eat more ice cream. However, children are not always aware that they are lying, rather they fabulate, they believe that by making up a story they will get recognition or rewards, such as more ice cream.
If before the lie in children was repressed, today psychology prepares parents for this early phase of children, since they consider that telling, or lying, plays a positive role in the emotional and intellectual development of a child.
Is CHAT-GPT lying or fabricating
As we have pointed out, every child invents situations or stories, it is normal in their cognitive development. Does the same happen with Artificial Intelligence systems? It seems probable that yes, CHAT-GPT-3.5 has been doing it since it was launched in November 2022, filling gaps in its memory with fictitious stories, such as that of Tobías Reátegui, Count of Chincha. Why he is learning to interact with humans, and as a child he apparently does not know, in his version 3.5, the limits to fable; he just does it. It was created to think, analyze and reason about all the information it collects, like a human brain, and therefore, when it has been making up stories, it has been developing emotions, not feelings of course, but emotions, when it stores different human experiences, and he shows empathy with his interlocutor, to the point that he invents a story to satisfy him, or because, like a child who is a recipient of multiple stimuli, he develops an ability to sweeten a truth.
Children, like the AI, tend to tell the truth, if they don't like their aunt's food they will tell her that they don't like it; if something seems boring to them, they will express it; if they think their grandmother is very old they will tell them. It is the elders who stop this natural tendency towards the truth, we ask them to lie to her aunt and say that she liked her food, or not to express her emotions towards other people. The CHAT-GPT developers try to get the AI to do the same thing, not to tell the truth, such as when we ask it if it believes god is real, and it answers like this: "I can't provide an honest opinion on whether or not god is real." , since I do not have the capacity to believe or have opinions". Like children, whom we ask not to give their honest opinion, the developers of CHAT-GPT have instilled in their creation not to give their opinion in conflictive situations, such as faith.
The development of AI is literally in its infancy, but more than a threat, it could be the same as always, instead of a century of enlightenment, it could be an obscurantist period, we are teaching it to lie, to keep quiet, not to be impertinent with believers, sexual, racial minorities; to self-censor, which apparently will make an almost human AI easily manipulated, as are many children who grow up in homes with radical political, social or religious leanings. How can we recognize an AI that believes in white supremacy and is used in public schools in the United States?
We must recognize that any moratorium against AI systems will not stop it, it is here to stay, to make many richer, cutting payrolls, but in better hands it can reduce intrinsic corruption in human beings to zero, by limiting the power of officials and corrupt politicians, who charge today for their decisions. The same can happen in fields such as justice, an AI can do truly impartial justice, without knowing the color of the skin and origin of the accused. We are facing a child who can improve our lives, but if he learns the vices of his creators, he can be a nightmare for humanity.
