Almost with total certainty you will have seen some type movie Terminator o Matix. Some in which artificial intelligence becomes aware of herself or something like that and creates murderous robots. Surely when thinking about the dangers of AI is one of the first scenarios that come to your head. However, people like Geoffrey Hinton do not agree with it.
And not precisely because he who is considered «the sponsor of AI» be optimistic with the future of this technology. It has been predicting a catastrophic future because of the chatgpt boom and all the alternatives that in recent months or years have seen the light. Hinton foresees a threat less similar to science fiction, but equally worrisome or more.
What is the greatest risk of AI? Geoffrey Hinton responds
He has done it during a recent interview. Since he abandoned Google, the truth is that Geoffrey Hinton has become the cricket of artificial intelligence. Rare is the week in which he does not perform any kind of public appearance to be strongly shown against it. But one thing is unquestionable: his speech at least always has coherence.
At 77, being a veteran of this world, Hinton It has been denouncing that it is necessary to regulate the AI if they do not want unpleasant surprises. But well, not how it is being done so far. Even so, it is not very optimistic about the legislation as God commands. He, like so many others, fears that the interest of prime companies above security.
In recent statements, the expert wanted to scare the fantasies of science fiction. The thing will not be so simple, he thinks, as murderous robots chasing the helpless human with their laser rays guns. Rather, the opposite will happen: Its main danger will come in the form of an alleged understanding «with a soft language». Manipulation, in a word.
«These are going to end up knowing much more than us,» says Geoffrey Hinton. «Actually, you know much more than us,» he says. «If you debated with them on any subject, you would get losing». And that can become a problem? For him, there is no doubt that yes. It will be as if the machines simply play with people.
Artificial intelligence, a potential manipulator
«Being smarter than us, they will be better manipulating people,» explains Hinton. His speech, in fact, even takes into account an aspect in which other specialists insist again and again: AI is not intelligent, nor does it have the most remote idea of what he is talking about. Simply simulates it, so to speak. In Geoffrey Hinton’s opinion, that is not a problem at all, on the contrary.
«He has learned with the example,» he explains, referring to what tools like Chatgpt have found on the Internet since they saw the light. «They have learned all those manipulative skills in a very simple way: just trying to predict the following word«He concludes.
Know How we work in NoticiasVE.
Tags: Artificial intelligence