The rise of artificial general intelligence — now seen as inevitable in Silicon Valley — will bring change that is “orders of magnitude” greater than anything the world has yet seen, observers say. But are we ready?
AGI — defined as artificial intelligence with human cognitive capabilities, as opposed to narrower artificial intelligence, such as the headline-grabbing ChatGPT — could free people from menial tasks and usher in a new era of creativity.
But such a historical paradigm shift could also threaten jobs and raise insurmountable social problems, experts warn.
Advances in technology from electricity to the internet have ignited powerful social change, says Siqi Chen, chief executive of San Francisco startup Runway.
“But what we are looking at now is intelligence itself… This is the first time we are able to create intelligence ourselves and increase its amount in the universe,” he told AFP.
Change, as a result, will be “orders of magnitude greater than any other technological change we’ve ever had in history.”
And such exciting, scary change is a “double-edged sword,” Chen said, envisioning using AGI to address climate change, for example, but also cautioning that it’s a tool we want to be as “controllable” as possible.
It was the release of ChatGPT late last year that brought the long-cherished idea of AGI one giant leap closer to reality.
OpenAI, the company behind the generative software that churns out essays, poems and computer code on command, this week released an even more powerful version of the technology that powers it — GPT-4.
It says the technology will not only be able to process text, but also images, and produce more complex content such as legal complaints or video games.
As such it “exhibits human-level performance” on some benchmarks, the company said.
– Goodbye to “hard work” –
The success of Microsoft-backed OpenAI has ignited an arsenal of sorts in Silicon Valley as tech giants look to push their generative AI tools to the next level — though they remain wary of chatbots off the rails.
Already, digital assistants from Microsoft and Google can summarize meetings, draft emails, create websites, create ad campaigns and more — giving us a glimpse of what AGI will be capable of in the future.
“We spend too much time consumed by the drudgery,” said Jared Spataro, Microsoft corporate vice president.
Through artificial intelligence Spataro wants to “rediscover the soul of work”, he said during a Microsoft presentation on Thursday.
Artificial intelligence can also reduce costs, some suggest.
British landscape architect Joe Perkins tweeted that he used GPT-4 for a coding project that a “very good” developer told him would cost 5,000 pounds ($6,000) and take two weeks.
“GPT-4 delivered the same in 3 hours, for $0.11,” he tweeted. “Really confusing.”
But that raises the question of the threat to human jobs, with entrepreneur Chen acknowledging that the technology could one day build a startup like his — or an even better version.
“How am I going to make a living and not be homeless?” he asked, adding that he was counting on solutions to emerge.
– Existential questions –
Ubiquitous artificial intelligence also puts a question mark over creative authenticity as songs, images, art and more are churned out by software instead of humans.
Will people eschew education, relying instead on software to do the thinking for them?
And, who is trusted to make the AI unbiased, accurate and adaptable to different countries and cultures?
AGI is “probably coming at us faster than we can process,” says Sharon Zhou, co-founder of a generative AI company.
The technology raises an existential question for humanity, she told AFP.
“If there is going to be something more powerful than us and smarter than us, what does that mean for us?” Zhou asked.
“And do we harness it? Or does it harness us?”
OpenAI says it plans to build AGI gradually with the goal of benefiting all of humanity, but it has admitted the software has security flaws.
Security is a “process,” OpenAI chief scientist Ilya Sutskever said in an interview with the MIT Technology Review, adding that it would be “very desirable” for companies to “come up with some kind of process that allows for slower releases of models with these completely unprecedented capabilities.”
But these days, Zhou says, slowing down just isn’t part of the vibe.
“The power is concentrated around those who can build these things. And they make the decisions about it, and they tend to move quickly,” she says.
The international order itself could be at risk, she suggests.
“The pressure between the United States and China was enormous,” Zhou says, adding that the race for artificial intelligence harkens back to the Cold War era.
“There’s definitely the risk with AGI that if one country figures it out faster, will they dominate?” she asks.
“And so I think the fear is, don’t stop because we can’t lose.”