AWith artificial intelligence, we have been experiencing for two years a transformation comparable in importance to the first industrial revolution. No offense to some, neither paradise nor the apocalypse is for tomorrow. We must expect many years of uncertainty.
AI remains limited to this day. She is not capable of cooking you a meal, walking your dog, much less waging a war or leading an organization. It’s not anytime soon that a malevolent superintelligence will take control of the planet. But progress is extremely rapid.
I see exciting advances in education, art, medicine, robotics… but I am also aware of the risks, most of which have not yet been taken into account. As was the case during the first industrial revolution, the effects on unemployment and income will be distributed unevenly, and often come without warning. There will be much suffering and worry, at all levels of society; as are new opportunities for many, and immense fortunes for some.
Human translators will be essential
The practical implications of AI can be surprising. Some examples. One day, an AI expert told me about the disappearance of human translators. Then another told me, on the contrary, that we were going to need it more than ever: AI allowing the rapid and inexpensive translation of “ absolutely everything”he continued, we will see an explosion in the amount of translated content, and we will therefore need human intervention to train and improve AI systems, as well as to review and correct the most difficult translations. important.
Read also | Article reserved for our subscribers Daron Acemoglu, Nobel Prize in Economics 2024: “The world must work for AI to serve humans”
Read later
This second point of view seems to me the most relevant. Especially since translation does not only concern weather reports or restaurant menus; it may also concern the FBI, the CIA, chemical companies, medical device manufacturers, emergency doctors, airplane pilots, special forces, or even suicide prevention hotlines! Human translators will be essential for training, monitoring and correcting AI systems.
You have 64.45% of this article left to read. The rest is reserved for subscribers.
