Recommended
Will artificial intelligence eventually completely surpass human capabilities? Will it replace humanity soon? Amid technological advances and the rapid escalation of AI skills, a group of specialists has predicted when both scenarios could occur.
A group of 2,778 researchers specializing in AI calculated when such a scenario could arrive for humans, and their estimates point to a year that is already generating debate: 2047. This is what the experts are saying.
According to research published in the scientific journal Journal of Artificial Intelligence Research, there is a 50% probability that machines could outperform human beings in all possible tasks by then.
Recommended
Ninel Conde presume su figura en redes, pero su abdomen desata una ola de críticas
Jugadores de los Knicks llaman a Taylor Swift la "fan más falsa" y desatan furia de las Swifties
¿Trabajas desde casa en EU? Te ofrecen $12,000 por dejar la ciudad y mudarte a las montañas
¿Vives en Estados Unidos y necesitas ayuda para pagar la luz y el gas? Estos inmigrantes pueden recibir hasta $1,000
¿Por qué Meghan Markle está distanciada de su padre? La historia detrás de su ruptura familiar
Although it does not mean that AI will take control of the world on that date, it does raise questions about the future of work, society, and the relationship between people and technology.
The study, titled Thousands of AI Authors on the Future of AI, gathered the opinions of researchers who had published work at major conferences and journals specializing in artificial intelligence.
The survey was conducted in October 2023, and its results show that technological development could advance faster than previously estimated. In fact, the 50% probability of reaching human-level performance in all tasks was moved up by 13 years compared to a similar survey conducted in 2022.
So… In what year could AI replace all human jobs?
The research also proposes a date for another scenario that could radically transform the economy: the complete automation of all human occupations. In this case, the researchers estimated that there is a 10% probability that it will be technically possible by 2037, while the probability reaches 50% only by the year 2116.
The difference between the two projections is important: just because a machine can outperform a person in any task does not necessarily mean that all jobs will disappear immediately, the scientists predict.
The concept of technological singularity is often used to describe a hypothetical moment in which artificial intelligence surpasses human capabilities and could develop new versions of itself at a speed that is difficult to control.
Futurist Ray Kurzweil has pointed to 2045 as an approximate date for that scenario. However, the survey of 2,778 researchers mainly measures the probabilities of reaching certain technological capabilities, not predicting an exact date for machines to rebel or take control of humanity.
Will AI be dangerous in the future?
The study also reveals concern about the consequences of an increasingly advanced AI. More than half of the participants considered that there are reasons to maintain considerable or extreme concern about risks such as misinformation, authoritarian control, and rising inequality.
Even among those who believe that superior artificial intelligence could generate more benefits than harm, some acknowledged that there is a possibility of catastrophic consequences, including human extinction.
[Publicidad]