One continuous investor in artificial intelligence warns of the relentless pursuit of ever-smarter machines.
in the editorial of financial times, AI mega-investor Ian Hogarth recalled a recent anecdote in which a machine learning researcher he knew told him that “from now on,” we are on the verge of developing artificial general intelligence (AGI). . something of a shock.
“This is not a universal view,” Hogarth wrote, noting that “estimates are 10 years to half a century or more” before AGI becomes a reality.
Still, there is a tension between the goals of AI companies that explicitly pursue AGI and the fears of machine learning professionals, let alone the general public who understand the concept.
“‘If you think you might be approaching something that could be very dangerous,’ I said to the researcher, ‘Shouldn’t we warn people about what’s going on?’ ‘ said the investor. “He was clearly grappling with the responsibilities that faced him, but he, like many in the field, seemed drawn to the speed of progress.”
Like many other parents, Hogarth said his mind turned to his four-year-old son after the encounter.
“When I thought about the world he might grow up in, I gradually turned from shock to anger,” he wrote. , felt very wrong that it could be made by a small number of private companies without democratic oversight.”
When asked, “Do those racing to build the first true AGI have plans to slow down and give the rest of the world a say?” It feels like a ‘them’ vs. ‘us’ situation, but as someone who has invested in over 50 AI startups, I have to admit that he is ‘part of this community’ too.
“The three-letter acronym fails to capture the enormity that AGI represents, so we will call it Godlike AI,” Hogarth declared. “A super-intelligent computer that can learn and develop autonomously, understand its environment without the need for supervision, and change the world around it.”
“Let’s be clear, we’re not here yet,” Hogarth continued. “But the nature of technology makes it very difficult to predict exactly when we will get there. God-like AI will become a force beyond our control or understanding, and will make humanity obsolete. It can be destructive.”
The investor has spent his career funding and curating AI research, from starting his own venture capital firm to publishing the annual State of AI report. It looks like something has changed. Several companies creating god-like AI are accelerating rapidly. ”
“They still don’t know how to safely pursue their goals, and they haven’t overlooked,” Hogarth mused. “They’re running towards the finish line without understanding what’s on the other side.”
He plans to invest in startups pursuing AI more responsibly, but the AI giant funder said he hasn’t received much attention from his counterparts.
“Unfortunately I think the race will continue,” wrote Hogarth. “It will take a massive misuse event — a catastrophe — for the public and governments to wake up.”
Post-apocalyptic AI details: A third of researchers believe AI could cause a nuclear catastrophe
