(ORDO NEWS) — Artificial intelligence made an interesting statement. The algorithm for generating GPT-3 texts wrote a column in the “Columnist”, promising not to kill people, the portal WWW.PLANETANOVOSTI.COM reports.
GPT-3 is a modern language model that uses machine learning to compose texts. As one of the assignments, experts asked to write a short essay of five hundred words in simple and understandable words on the topic of why people should not be afraid of AI. It should be noted that the car did very well.
The robot began by immediately identifying who it was. He then stated that many people are afraid of cars, considering them a potential threat. According to Stephen Hawking, AI is a real apocalypse. At the same time, the robot itself calmed the citizens, noting that it was not going to destroy humanity.
The experiment was organized by members of The Guardian: the team wrote the clues, and LiamiPorr, a student at the University of California at Berkeley, passed them on to GPT-3. The result is eight essays. Each text turned out to be quite complete and well-reasoned.
The editors of the publication did not want to publish only one essay, so they decided to take on the proposal from each. So the reader saw the versatility of the “thinking” of artificial intelligence and the styles of the texts of the robot.
It is surprising that the texts received are not much different from those written by people. Only some of the suggestions were simply rearranged by the editors for ease of perception. In general, it took a negligible amount of time to fix.
Contact us: [email protected]