One of the dangers that has folks worried about the heavy use of AI is the possibility that our cognitive skills will atrophy from lack of use. Wall Street Journal tech reporter Sam Schechner recently wrote an essay describing how reliance on ChatGPT had eroded his skills (here’s a link to a non-paywalled article about Schechner’s essay). He’s not alone in this concern.

One of the most consistent fears I hear from audiences around the world is that over-reliance on AI will degrade our human mental capacity, and the worry is greatest among parents who fear for AI’s impact on their children.

The danger of AI-abetted mental atrophy is real. But it is not new, and it is solvable.

I am old enough to remember the panic people felt about how calculators and computers would stunt human mathematical ability. Or later, how the internet would make plagiarism inevitable and stunt our writing ability. Each of these fears had some basis in reality, but we quickly found ways to address them.

In mathematics, we teach basic math without a calculator, and then incorporate calculators into more advanced courses. In writing, we ask students to write timed, in-person essays so we can assess their ability to compose coherent arguments unaided by other means.

I’m writing this essay on a computer, which is the tool I use for nearly all my waking hours and nearly all my work. It hasn’t atrophied my mind; it has allowed me to expand its capability far beyond its flesh-and-blood limits.

But perhaps you fear that AI is different; that it fundamentally changes the nature of mental work, and thus the examples I provide above don’t apply. If that’s the case, then let’s consider the impact of technology on human physical capabilities.

The subtle but revolutionary power of the Industrial Revolution is that it allowed us to broadly substitute mechnical power (steam and electricity) for muscle power. This massively increased productivity, but it also changed the nature of our daily lives.

Where before we had toiled in the fields or worked with our muscles all day to forge horseshoes or weave cloth, we suddenly shifted to working in factories, and later, desks. The result was a decline in physical fitness. Muscular atrophy, if you will.

One possible response to this decline would be to abandon industrial civilization and go live in the wilderness. This might make for good literature, but it is remarkably unappealing to the vast majority of people. Industrial society needed to find a more effective solution.

The answer we eventually settled on was deliberate physical exercise. Rather than developing our physical capabilities as a by-product of 8+ hours of daily work, we found ways to exercise our muscles in a small fraction of the time. Even though this deliberate exercise didn’t produce valuable goods or services, we were still better off economically by working our industrialized jobs, and then investing some of our leisure time into exercise. By the 20th century, exercise was part of our school systems in the form of “physical education” and gymnasiums and health clubs like the YMCA had sprung up to meet the needs of the adult population.

This isn’t a perfect solution; many people don’t bother to exercise, and live unhealthy, sedentary lifestyles. But almost everyone has the option to exercise, even if just in the form of simple calisthenics like jumping jacks.

And as we continue to develop the science of exercise, the upper bounds of human ability continue to increase. New athletic world records are set every year, and today’s athletes are bigger, stronger, and faster than at any other time in history. An athlete like LeBron James would seem literally superhuman to a 19th century athlete, let alone a British factory worker.

I believe we can do the same thing for deliberate cognitive exercise that the reformers of the past did for deliberate physical exercise.

Like the workers of the pre-industrial past, right now we largely rely on our daily work for our post-school cognitive exercise. If the widespread adoption of AI causes us to lose this source of incidental cognitive exercise, we need to find a way to make up for it with deliberate cognitive exercise.

And just like manual labor is far from the most effective form of physical exercise, attending Zoom meetings and writing weekly reports is far from the most effective form of cognitive exercise. We can and will develop new forms of deliberate cognitive exercise (likely with the aid of AI) that will train our mental capabilities in a fraction of the time. Perhaps we will even be able to develop certain cognitive abilities to a greater degree than our ancestors, continuing to drive the Flynn effect for subsequent generations.

But just as many in our industrial society don’t make the time for physical exercise, many won’t make the time for cognitive exercise either. Whether your mental abilities atrophy or develop will be a result of the choices you make. You could become like a mental version of LeBron James, performing a high level well beyond our traditional concept of our prime years, or you could allow yourself to become like a mental version of Wall-E’s chairbound passengers, utterly dependent on the machine intelligence around you.

Which will you choose? Extra mental stimulation? Or the latest Netflix binge?