A couple of weeks ago, Nobel laureate Demis Hassabis forecasted that artificial intelligence (AI) will match human intelligence within the next 5–10 years. This expectation, shared by other AI analysts, is based on the exponential growth in AI technology. However, it neglects an important factor. As humans delegate thinking to computers, they get dumber.
The situation is similar to athletes losing muscle tissue by not practicing. The introduction of AI to the workforce relieves humans from a variety of intellectual tasks. The resulting “intellectual vacation mode” is likely to trigger a decline in human cognitive capabilities. As a result, the crossing point between human intelligence and AI might arrive faster than expected, not only because AI abilities are advancing fast but also because humans are getting dumber. The two trends are correlated since better AI agents would reduce the intellectual burden on humans.
A paper published in the journal Intelligence two years ago concluded that American IQs are dropping for the first time in recent history. An analysis of 394,378 scores on intelligence tests taken between 2006 and 2018 showed that IQs had fallen in the categories of logic and vocabulary, computation and mathematics, and visual problem-solving and analogies. This reverses the so-called Flynn effect, which previously described a steady rise in IQ by about three IQ points per decade since 1932. The decline is steepest in the age group of 18 to 22, which is supposed to have better IQ scores than older people. Separate from that, a 2023 survey revealed that 34% of American adults scored at the lowest levels of numeracy, lacking the ability to work with numbers.
Another recent study reported a trend that started around the mid-2010s with more people finding it difficult to concentrate, reason and problem-solve. Young adults exhibit a reduced attention span and weakening critical thinking skills based on data concerning 18-year-old Americans as well as data on 15-year-olds around the world. Extensive exposure to digital screens was demonstrated to cause a decline in verbal skills of children and reduce the ability of college-age adults to concentrate and process information.
A psychological study revealed that people could retain and process data significantly better if their smartphones were in another room. Just turning their phone off did not make a difference, as people still demonstrated a decline in cognitive abilities when their smartphone was next to them.
Human intelligence flourishes in response to challenges. With fewer problems to solve and more sources of superficial engagements in the age of social media and AI, the need for deep thinking and rigorous inquiry declines.
Of course, there will always be statistical outliers at the high-IQ tail of the probability distribution, involving people who are driven to learn, analyze, and innovate without the help of AI agents.
Given these trends, one may wonder optimistically whether the exponential rise in AI will compensate for the alarming decline in human cognitive abilities, so that the net effect will still be positive. I do not think so.
The prosperity of democratic countries depends critically on the cognitive abilities of their citizens and less so on the tools they use. This is because humans control the physical world in ways that cannot be substituted by AI systems. We need politicians to make sound judgments, juries to reach reasonable verdicts, technologists to make good investments and scientists to discover new knowledge. But more than anything, we need the commitment of humans to interactions with other humans and to the challenge of finding the truth from primary sources of evidence and knowledge. Consuming knowledge that was processed by AI is equivalent to feasting on junk food. It may feel good at the moment, but in the long-term it is bad for our health.
Here’s hoping that we can revise our education system so as to reestablish a rising Flynn curve of human IQ scores in the 21st century. Otherwise, we might not be able to get a hold of the steering wheel when our self-driven cars will ride us off the cliff.
ABOUT THE AUTHOR
Avi Loeb is the head of the Galileo Project, founding director of Harvard University’s — Black Hole Initiative, director of the Institute for Theory and Computation at the Harvard-Smithsonian Center for Astrophysics, and the former chair of the astronomy department at Harvard University (2011–2020). He is a former member of the President’s Council of Advisors on Science and Technology and a former chair of the Board on Physics and Astronomy of the National Academies. He is the bestselling author of “Extraterrestrial: The First Sign of Intelligent Life Beyond Earth” and a co-author of the textbook “Life in the Cosmos”, both published in 2021. The paperback edition of his new book, titled “Interstellar”, was published in August 2024.
