Could humans suffer the same fate as dinosaurs?
Although there is still much uncertainty about the potential risks posed by artificial intelligence, some experts are concerned that AI could cause the extinction of humanity.
According to a paper for whom 2,778 AI researchers were surveyed; A majority of scientists (58%) believe there is a 5% chance that AI will wipe out humanity or have some other catastrophic outcome.
“This is an important sign that most AI researchers don't think it's highly implausible that advanced AI will destroy humanity,” said co-author of the paper, a Berkeley, Calif., machine. said Katja Grace of the Intelligence Institute. new scientist. “I think the general idea that the risk is not trivial says much more than the exact percentage of risk.”
Respondents said the technology had a roughly 50 chance of successfully completing a specific sample task (such as writing a song to rival Taylor Swift's work or coding a payment processing site from scratch) within 10 years. % or more.
But more complex tasks, such as installing electrical wiring or solving mathematical puzzles, are expected to take much longer.
It is estimated that there is a 50% chance that AI will outperform humans at every task by 2047, and a 50% chance that all human jobs will be fully automated by 2116. However, both estimates are older than those presented in his 2022 edition of the same study.
The study is the largest of its kind to date, and provides researchers with insight into what the timeline for future AI advancements and milestones will look like, as well as the impact AI will have on society, for better or worse. We asked for opinions on the matter.
The authors cautioned that AI researchers are not experts in artificial intelligence prediction.
Additionally, this type of research “doesn't have a good track record” of predicting AI outcomes, so there's no immediate cause for concern, Emil Torres of Case Western Reserve University in Cleveland told the magazine.
In fact, previous research has shown that predictions made by AI experts are no more accurate in the long run than those made by the general public.
“Many of these breakthroughs are completely unpredictable,” Torres said. “And there's a good chance the AI field will experience another winter.”
Meanwhile, researchers had more pressing concerns about AI.
More than 70% of those surveyed expressed serious or extreme concerns about scenarios such as deepfakes, public opinion manipulation, engineered weapons, authoritarian control, worsening economic inequality, the spread of disinformation, and deteriorating democratic governance. expressed concern.
“We already have technologies that can seriously harm society, right here, right now. [the US] It’s a democracy,” Torres said. “Let's see what happens in the 2024 election.”
