Researchers from New York University surveyed professionals in the field of Natural Language Processing to gather opinions on the current state of AI and what the types of ethical concerns they may have for it in the future.
Not only did the respondents conclude that AI could overturn our current society, they also predicted that its mishandling could result in something as catastrophic as nuclear war. And perhaps the most interesting part of this is the fact that the survey was conducted last May — before the infamous ChatGPT was even in the public eye.
By the Numbers
Out of 480 researchers polled, a full 73% said that AI as an automation in the workforce could lead to “revolutionary societal change” during this century. They compared its impact as that of the Industrial Revolution when steam power and communications advancements completely changed the world.
As far as the nuclear war predictions, 36% of those surveyed agreed that AI has the power to cause “catastrophic outcomes” that could lead to an “all-out nuclear war.” While there were no further details given by those polled, we can’t help but wonder what they might say now that AI has taken center stage in recent months.
— Terror Alarm (@Terror_Alarm) February 27, 2023
We’ve already seen questionable behavior from AI in early public trials. Everything from passive-aggressive temper tantrums to stalker-like tendencies has caused AI programmers to pull in the reins a bit, but we can’t help but wonder if those polled in last May’s survey may be on to something.
Elon Musk, who has never been one to hold back on his opinions, has famously said that AI is “one of the biggest risks to the future of civilization.” That should probably scare us considering he may know more of the behind-the-scenes stuff than most people.
Former Google CEO Eric Schmidt also considers the warfare side of the AI coin as he was quoted as saying that he believes AI could have as big of an impact on combat as nuclear weapons did.
“Einstein wrote a letter to Roosevelt in the 1930s saying that there is this new technology — nuclear weapons — that could change war, which it clearly did,” Schmidt told Wired in a February interview. “I would argue that [AI-powered] autonomy and decentralized, distributed systems are that powerful.”