The rapid advances in Artificial Intelligence (IA) have raised many reservations. Apparently, even Geoffrey Hinton, considered the “godfather of AI” is alarmed about the risks of this technology. For the computer scientist and cognitive psychologist, Big companies' aggressive race to develop new solutions is getting dangerous.
Recently, Hinton is no longer part of the Google team and declares that he regrets his dedicated work in developing artificial intelligence.. Part of this would be related to the possibilities of cybercrime, but, also the use of the tool for suspicious or unnecessary purposes.
“It’s hard to see how to stop bad actors from using this to do bad things.”, these Hinton. The big question is that, as they have a very vast intelligence, These artificial solutions can also hide hitherto unknown dangers for humanity.
Para a BBC, the expert stated that chatbots (virtual robots) pose quite frightening risks. “Right now, they are not smarter than us, as far as I know. But I think they could soon be. What we're seeing is things like GPT-4 surpassing a person in the amount of general knowledge they have, and surpasses it by far”, evaluate.
Hinton also states that digital systems work very differently compared to human capabilities. “It's like you have 10 thousand people, and whenever one of them learned something, everyone would automatically learn. And that's how these chatbots are able to know much more than anyone.”
Hinton's achievements
Regarding the dismissal from Google, the scientist said that his age would be the big question, since today there is 75 years. Hinton has been working in the research field for 50 years and, em 2012, the mathematical system of neural networks, foundation of machine learning technologies, was developed by him and two of his students.
After that, the scientist and his team were hired by Google, where he worked for more than a decade. For him, Artificial learning models have always had lower capacity than human knowledge. Although, as AI evolves, the scientist began to fear that technology would replace human capacity.
Experts sign public letter
Rapid adoption of AI tools, ChatGPT e Midjourney, has worried experts in recent semesters. That's why, involved in the segment seek to develop new security protocols to deal with this technology that still raises many doubts.
In this last month, around a thousand of them signed a public letter requesting a pause in algorithm training. Soon after, the Association for the Advancement of Artificial Intelligence also released a document that warns about potential risks of AI.
