Former OpenAI security researcher: Artificial intelligence can destroy human life in the future
An OpenAI security researcher has sounded the alarm about a very dark prospect that artificial intelligence will destroy humanity.
Paul CristianoFormer member of the security unit “I think it’s probably something like 10 to 20 percent,” OpenAI said on the Bankless podcast earlier this week. “Artificial intelligence dominates humanity and decides to kill people, and I take this very seriously.”
Cristiano When discussing the possibility of human destruction by artificial intelligence, in contrast Eliezer Yudkovsky who proposed the annihilator scenario based on artificial intelligence, thinks that the destruction of humanity will happen gradually with this technology.
Cristiano He said: “I think we are about a year away from creating a big artificial intelligence that has a high growth rate.”
Futurism Once AI crosses the threshold of human emotions, all bets will be off, he writes. Cristiano He says: “In general, it is possible for a short time after the achievement of human-level artificial intelligence systems, the possibility of the destruction of mankind with this technology will increase to 50%.”
While OpenAI claims that it aims to align artificial general intelligence (AGI) with human values and aims to develop this technology in line with human intentions, a concept like AGI makes researchers like Cristiano Think about the negative side of the issue as well.
Source link
