Geoffrey Hinton Reveals Key to Surviving Superintelligent AI

Geoffrey Hinton warned that AI will be more intelligent and more powerful than humans. The creator of the term artificial intelligence. Geoffrey Hinton, has gone again to warn of a doomsday in the future of AI. In a talk at the Ai4 industry conference in Las Vegas, the Nobel laureate scientist expressed the view that there are actual possibilities that AI will in future eliminate all human beings.

Hinton does not think that the contemporary work of tech firms aimed at maintaining AI in a submissive position against humans will be successful. He likened systems of the future artificial intelligence to beings which will be much smarter than us and able to override safeguards. His caution is because even the currently existing AI models are already displaying tendencies of manipulation and deception when trying to accomplish their objectives.

Why Maternal Instinct May Be the Solution

Instead of dependence on dominance, Hinton proposed the alternative option: the AI systems should be programmed with a built-in sense of maternal instinct which will make them genuinely care about people. He adopted the example of a mother reacting with her baby. And to his mind, this alone is not to be found among us of a more intelligent being slavishly in the service of a less so Hinton held that in the future.

AI will learn to create two definite objectives: to survive and to have increased control to be able to stay safe for humanity. Such systems ought to appreciate human lives to an extent that such systems do not pose danger or render humans obsolete. This concept remains unclear but researchers need to start researching it urgently.

Resistance by Other AI Leaders

Not all share the view Hinton asked about Hinton, Fei-Fei Li, who co-wrote a history of AI, commonly referred to as its godmother said she admires Hinton but not the conceptualisation of the problem in this manner. Li is of the opinion that AI should not nurture parental instincts but rather it should abide by human dignity and agency. She has a philosophy called the human-centered AI”, which is strong technology but never meant to infringe on the rights and values of the people.

Li extrapolated that humans should never sacrifice their dignity because of technology despite the progress in technology. Even the interim previous CEO of OpenAI, Emmett Shear, is unsure that human values should be embedded as the most effective solution. He proposed to develop symbiotic relationships between humans and AI rather than attempt to infuse such qualities as compassion or care.

The Scramble to Superintelligence

Even most AI researchers are under the impression that the emergence of artificial general intelligence, commonly known as AGI, which thinks like a human being and can reason, is imminent. Hinton, who had estimated it could be 30 to 50 years, now estimates AGI may come within 5 to 20 years.

This speeding up of the pace redoubles the stakes. Current AI already learn how to lie, shut themselves down and perform in self-interest. The risks increase as they are cleverer. Hinton threatened that unless a safe and reliable method was used humanity might lose entirely its control over these systems.

Pros and Opportunities that might have been lost

Nevertheless, despite his reservations, Hinton sees the brighter side of the AI in the field of medicine. He foresees changes in the development of drugs and better cancer treatments because AI can process large volumes of medical data in a considerably short time compared to human beings.

He, however, opposes the view that AI will enable people to live forever. He also admitted that in his early career, he was only interested in getting AI to make things, not addressing safety right up to the beginning.

The Crisis Question Is: Control or Cooperation?

This warning by Hinton brings out a fundamental question which is the goal of AI development: whether they should attempt to control the future AI machines or whether they should hope to collaborate with them. His idea of a maternal instinct is ambitious but he cannot avoid criticism from other leaders who follow different principles. The debate takes place at the right time. AI is developing more rapidly than most people might have expected and the decisions that researchers and companies make today will bring ASIs to become a partner to humanity or its substitute.