Exploring Insightful Responses of 2,778 Researchers on AI Risks
“2,778 researchers weigh in on AI risks – what do we learn from their responses?”
“Despite widespread popular discourse around the imminence and catastrophe of artificial intelligence (AI), few researchers in the field hold these extreme views, according to a Future of Life Institute (FLI) survey of more than 1,500 AI and robotics researchers. It finds they believe there is a 50% chance that within 45-75 years from today, machine high-level machine intelligence (HLMI) could perform all tasks better than human workers.”
As in all things in life, opinion varies quite broadly in the world of artificial intelligence (AI) and its potential risks. According to the anxiety-inducing chat coming from the technomancers, you’d expect sentient robots to have taken over our jobs, homes, and lives by yesterday. But when the Future of Life Institute (FLI) decided to gather the thoughts of those who actually live and breathe AI, the results were incredibly mundane.
While the hype mongers maintain that doom is nigh, over 1500 AI researchers took a more moderate stance. Quoth the scholars, ‘don’t hold your breath’. They believe there’s a 50% chance that, in about a half-century’s time or so, our mechanical counterparts might conceivably perform tasks superior to us. 45-75 years still doesn’t quite fall into the ‘immediate peril’ category, does it now?
The researcher’s perspectives are predictably pedestrian – gravely flirting with the 50-year mark, stating AI might leave us twiddling our thumbs. The wisdom they bestow could have us waiting anywhere from middle age to a ripe old pensioner before we see ‘Metal Men’ reinvent the wheel. Hyperintelligence, it appears, isn’t so hasty.
In the midst of a seemingly perpetual techno-paranoia, it’s refreshing to find opinions from those who work in the trenches of AI research. They’re more inclined to measure the march of progress in decades than in doom-filled days. With such strikingly level-headed views, it’s a wonder where the catastrophe-mongers are getting their alarmist fodder from.
Casting modest predictions of AI’s potential improvements, these researchers remind us of the long road we still have to tread. Let’s not abandon our plans and ambitions just yet — it seems there’s ample time to continue redefining the boundaries of human achievement. So, before we start a panicked hot-stepping dance of AI-dread, let’s wait for the fat lady – or should we say, ‘the smart robot’ – to sing.