Automation researchers believe that there is 50% chance that within 45 years, artificial intelligence makes it possible to carry out all the tasks currently made by men.
Must wait 120 years, so that the probability that the set of jobs currently occupied by men are fully automated, i.e. 50%.
The first tasks to be impacted through automation will be as follows: transcribe a speech (within 8 years), translate a text (within 8 years), or advise clients on the phone (within 9 years). The profession of surgeon could be automated within 37 years.
Nearly half of experts in the field believe that work automation will have a positive or even very positive impact on our daily lives, according to one study produced by Yale and Oxford. The two prestigious universities jointly interviewed 352 researchers in automation and machine learning, and asked them from when they estimate that robots can completely replace us.
And the answer is clear: there is a 50% chance that within 45 years, a robot will be able to perform all the tasks that we know how to do today. In Asia, such a probability is reached within 30 years, while in North America, it will take almost 75 years to reach it. Europe is in the world average.
On the other hand, it will take 122 years for there to be a one in two chance that all the tasks we perform today will be performed by robots, better and at lower costs.
If we consider more specifically the steps to be taken to achieve this total automation of work, the first tasks likely to be automated will be the transcription of a speech (within 8 years), or its translation (within 8 years). years also). It will take about ten years for robots to write an essay, and 11 years for them to be able to drive a truck. Within 15 years, they should be able to replace an in-store salesperson, write a bestseller for the New York Times within 35 years, and perform surgery within 37 years. We regret that the study does not seek to determine what will then be the place of man in this world where work will be “totally automated”.
** Study carried out by Yale and Oxford, on a sample of 352 researchers in automation and machine learning.