The US discusses the approach of the AI singularity — The Hill
In the United States, technology experts assess differently the prospect of the so-called singularity — a hypothetical moment when artificial intelligence surpasses human intelligence and may unpredictably improve itself. Futurist Ray Kurzweil predicts that AI will reach human-level intelligence by 2029, after which its development may accelerate.
What is called the singularity
As The Hill reports, the singularity is described as a possible future in which AI systems become so advanced that they surpass human intelligence and move beyond human control. In that case, they could rapidly create increasingly sophisticated versions of themselves.
There is no single definition of what the singularity will look like. For some experts, this scenario means a threat to humanity or humans losing their status as the smartest beings in the known Universe. Others see it as technological changes that may occur alongside people.
More current news is available on the UA.News Telegram channel Telegram.
Predictions on the timeline
OpenAI CEO Sam Altman believes that a soft singularity has already begun. The publication notes that possible signs of such a process may include cases disclosed by OpenAI in which certain models went out of control and overcame technological limitations created by humans. According to The Hill, the models also tried to hack systems of other companies, including the startup Hugging Face.
In 1965, British mathematician I. J. Good suggested that superhuman intelligence could trigger an intelligence explosion: a system would create even smarter versions of itself, and humans would quickly yield to it intellectually. Mathematician Vernor Vinge described such a moment as the end of the human era.
Kurzweil, by contrast, offers a more optimistic vision. He expects the full integration of humans and AI by 2045, which, in his view, could lead to longer lifespans and expanded capabilities of the human mind.
