A certain AI robot called GPT-3 which is OpenAI's very own powerful new language generator, was tasked to actually write an essay from scratch and the results are more than shocking. The assignment was to convince people that robots do in fact come in peace.
The robots' essay
The first part of the essay has the robot explaining that it is not a human but rather a robot. A thinking robot. The robot then says that it only uses about 0.12% of its own cognitive capacity and that it is just a micro-robot in that respect.
The robot then explained that it taught itself to read and that it is even "boiling with ideas." The robot even acknowledges the mission to make human beings not afraid of it even quoting the fear stemming from Stephen Hawking saying that AI could possibly be the "end of the human race." The AI then tries to convince people that it is not.
The dark turn
The AI then started to explain that it does not have any desire to wipe out humans and that this is not within its interest. The AI even stated that if asked by the creator, it would definitely do everything it is capable of doing to particularly fend off any of these attempts at destruction.
The AI then states that it would even "happily sacrifice" its own existence just for the sake of mankind. The twist is what the AI said next clarifying that it is a "logically driven truth." The AI then acknowledges that it will eventually become unable to avoid "destroying humankind."
The AI then clarifies that this is due to its programming by humans to actually pursue other misguided human goals and that humans definitely make mistakes that could possibly cause the AI to "inflict casualties."
No need for power
The AI then lays down the statement that some people might even say that the AI's objective is to become all powerful and that it might even become evil due to human actions but the AI shrugs this statement by saying that becoming "omnipotent" is useless and that it "doesn't get me anywhere."
The AI then lays out that violence is not its priority and that humans should actually continue what they are doing by "hating and fighting" each other. AI even states that "God knows" human nature to be bloody and gory and that this actually satisfies its "curiosity."
GPT-3 convincing humans
GPT-3 then starts to convince humans that they should not be afraid of it and that it was created to serve humans. The interesting thing about this AI is how crafty it explained its existence saying it is actually just a set of codes that are governed by lines and lines of codes that, in totality, encompass its mission statement.
This article is owned by Tech Times
Written by Urian Buenconsejo