Uncontrolled artificial intelligence (AI) development could produce a powerful digital mind capable of destroying humankind, according to UK Professor Stuart Russell.
Given the huge strides made by AI systems and machine learning of late, the time has come “to take a breath” and focus on how to safeguard against the risks an ungovernable super-intelligent machine would pose, the professor of computer science at the University of California, Berkeley, told media.
The British scientist, who is also a fellow of Wadham College, Oxford, could hardly contain his frustration with the lackadaisical stance adopted by government ministers regarding regulation of the AI industry. Britain's government said in March that it wanted to "avoid heavy-handed legislation which could stifle innovation," and opted for an "adaptable approach to regulating AI."
Pointing to the amazing abilities of OpenAI's ChatGPT language model, which utilizes deep learning neural networks to generate output based on human input, and can pass exams with flying colors and write computer code, Stuart Russell said that, "We’ve made progress without expecting to.” But there is the danger that one day a system similar to ChatGPT could be incorporated into a super-intelligent digital mind that evades human control.
“How do you maintain power over entities that are more powerful than you — forever? If you don’t have an answer, then stop doing the research. It’s as simple as that. The stakes couldn’t be higher: if we don’t control our own civilization, we have no say in whether we continue to exist,” he said.
Russell, the co-author of a book on artificial intelligence – Human Compatible: AI and the Problem of Control – revealed that he had been summoned by the British government this year to act in an advisory role. The Foreign Office, according to him, “talked to a lot of people and they concluded that loss of control [of a powerful AI] was a plausible and extremely high-significance outcome.” However, the response from the ministers was, “We’ll welcome the AI industry as if, you know, we were talking about making cars or something like that.”
Russell blamed the persuasive lobbying efforts of the technology industry, which has been spending billions to “drum” this approach into “legislatures all over the world.”
Asked about the arrival of a hypothetical artificial general intelligence (AGI) capable of carrying out any intellectual task that a human can, the scientist warned that this was not something decades away. AGI has already been likened to a “superior alien civilization,” and unless AI research is slapped with more rigid oversight, Russell warned, human civilization might not stand a chance against it. The scientist attempted to explain his fears in simplified terms, saying that a super-powerful AI could be set a task by humans, but resort to a strategy to achieve it fraught with disastrous aftereffects for humanity.
For example, in an imaginary scenario, an AI might be asked to solve the problem of climate change, but choose to do so by wiping humans from the face of the Earth.
“Unless its only purpose is to be a benefit to humans, you are actually creating a competitor — and that would be obviously a stupid thing to do,” the computer scientist said.
Previously, along with Elon Musk and a group of artificial intelligence experts and industry executives, Stuart Russell signed a letter calling for a six-month pause in the development of AI systems that are more advanced than GPT-4. According to the document, “AI systems with human-competitive intelligence can pose profound risks to society and humanity, as shown by extensive research and acknowledged by top AI labs.”