A new Yahoo/YouGov survey reveals a deep undercurrent of fear among Americans regarding the future of artificial intelligence. Despite Silicon Valley’s excitement and Wall Street’s investment boom, over half of Americans—53 percent—now believe it is likely that AI will eventually “destroy humanity.”
The poll, conducted in late October among 1,770 U.S. adults, comes amid major AI advancements, including the rise of tools like OpenAI’s ChatGPT Atlas and Nvidia’s record-breaking $5 trillion market cap. While 85 percent of those who’ve used AI chatbots say they’re at least somewhat helpful, the optimism ends there.
More than half of respondents (59 percent) said they encounter AI-generated content online too often, and 56 percent admitted they’ve mistaken it for real content. Though 54 percent feel confident in their own ability to distinguish AI-generated content from real content, only 21 percent believe the average person can do the same.
When asked how AI impacts personal intelligence, only 16 percent said AI makes them smarter. Thirty percent believe it helps in some areas but makes them “dumber” in others. Just 17 percent see AI having a mostly positive effect on their lives, while nearly twice as many (32 percent) expect a negative impact.
The poll also presented respondents with predictions about AI’s future. Sixty percent believed AI will eventually help solve complex problems like curing cancer. But even more—63 percent—said they believe AI will become so advanced that humans will lose control over it. Most alarming: 53 percent said it’s “somewhat” or “very” likely that AI will destroy humanity.
Even when asked about AI’s greatest promise, like curing cancer, only 15 percent said that outcome was “very likely.” In contrast, 20 percent said the total destruction of humanity by AI was “very likely.”
