The debut of artificial intelligence systems with capabilities on par with the human brain could be just a few years away, according to the boss of the AI lab Google DeepMind.
Google DeepMind CEO Demis Hassabis touted rapid gains in the development of so-called “artificial general intelligence,” which he defined as an advanced AI system with “human-level cognitive abilities.”
“The progress in the last few years has been pretty incredible,” Hassabis said during a Tuesday appearance at the Wall Street Journal’s Future of Everything Festival.
“I don’t see any reason why that progress is going to slow down. I think it may even accelerate,” he added.
“So I think we could be just a few years, maybe within a decade away.”
The development of artificial general intelligence (AGI) systems is subject of heated debate and competition in the AI sector – with industry leaders weighing its potential benefits against mounting warnings of the risks it could pose to society.
Hassabis’ bullish remarks on AI came just days after Dr. Geoffrey Hinton, a renowned computer scientist, quit his job at Google so that he could speak more freely about his concerns regarding unrestrained AI development.
Hinton, who is known as the “Godfather of AI,” said advanced systems are progressing at a “scary” pace — and warned that bad actors will likely exploit them “for bad things” that will be hard to prevent.
The 75-year-old added that he now partly regrets his life’s work due to the technology’s risks.
Google DeepMind and Microsoft-backed OpenAI are among the firms who have expressed interest in creating AGI systems.
Hassabis said that AI researchers could develop “very capable, very general systems” with human-like intelligence within “the next few years.”
“The brain is the only example we have of a general intelligence in the universe, as far as we know,” he said.
OpenAI CEO Sam Altman addressed the pros and cons of a theoretical human-level AI system in a lengthy blog post in February.
Altman noted its potential benefits include “[elevating] humanity by increasing abundance, turbocharging the global economy, and aiding in the discovery of new scientific knowledge that changes the limits of possibility.”
At the same time, the technology would carry a “serious risk of misuse, drastic accidents, and societal disruption,” Altman warned.
“Because the upside of AGI is so great, we do not believe it is possible or desirable for society to stop its development forever; instead, society and the developers of AGI have to figure out how to get it right,” he said.
Alphabet has ramped up its AI efforts as it scrambles to keep pace with rival Microsoft, which is backing OpenAI.
Hassabis was named the CEO of Google DeepMind last month after the company combined its two main AI initiatives, Deepmind and Brain, into one unit.
In a widely scrutinized interview with “60 Minutes” last month, Google CEO Sundar Pichai warned that society would need to adjust to rapid advancements in AI that could shake up the job market and cause other worries.
Pichai also discussed strange happenings in which Google’s AI programs have developed “emergent properties,” or learned skills it was never trained to know. In one case, a Google program taught itself to translate the language Bengali.
“There is an aspect of this which we call, all of us in the field call it as a ‘black box,’” Pichai said.
“You know, you don’t fully understand. And you can’t quite tell why it said this, or why it got [it] wrong.”