As investment in artificial intelligence surges, brain health and cognitive skills are deteriorating. The future will depend not only on better machines, but on rebuilding the human intelligence needed to use them.
AI has spread faster than any technology in modern history. Within three years of ChatGPT’s release, generative AI has been adopted by 53 per cent of the world’s population – faster than the PC and the internet – while organisational adoption hit 88 per cent last year. Global corporate investment in AI has more than doubled, with much of that money going into physical infrastructure like data centres and power grids.
Scepticism about AI is rising almost as quickly. A survey by S&P Global found that 42 per cent of companies abandoned most of their AI initiatives in 2025, up from 17 per cent the year before, and in a recent PwC poll, 56 per cent of CEOs reported that AI has brought their firms no measurable benefit. Communities around the world are now mobilising to prevent data centres from being built in their backyards, while documented incidents of AI harms rose by more than 55 per cent last year.
Yet even as governments rewrite their industrial strategies around a technology whose value they cannot yet quantify and investors, insisting that this time is different, drive valuations reminiscent of the dot-com boom, the calls for more responsible and ethical AI are growing louder. Everyone is trying to divine the future of AI in the hope of mastering whatever comes next.
The utopian and dystopian camps suffer from the same myopia. Both assume that the future of humanity depends primarily, perhaps exclusively, on the future of AI. It does not. And while optimists and pessimists alike have trained their focus on the rise of AI, they have ignored the rapid decline in brain health and brain skills.
That decline is eroding the human capacity on which every system, including AI, depends. Mental and neurological conditions, substance-use disorders, strokes and self-harm now account for roughly one-quarter of the global disease burden. The number of adults over 65 living with dementia rose from 18.7 million in 1991 to 49 million in 2021 and continues to climb. Governments allocate just 2 per cent of their health budgets to mental health, even after the World Health Organization declared loneliness a “pressing health threat” in 2023.
Brain skills are in freefall too. A majority of business leaders surveyed in 2025 by the skills training company General Assembly said entry-level employees were less prepared for their jobs than they had been five years ago, especially when it came to soft skills. Likewise, the OECD’s 2024 Survey of Adult Skills, the largest direct assessment of adult cognitive proficiency ever conducted, recorded declining or stagnant literacy across most participating countries. Roughly 18 per cent of adults in OECD countries lack basic proficiency in literacy, numeracy and problem-solving. Even among university-educated adults, proficiency fell or flatlined in most of the countries surveyed. New data from the Programme for International Student Assessment show that student performance in reading, mathematics and science is declining across the OECD.
AI is fuelling these trends. Emerging research suggests AI can reduce cognitive engagement, with studies finding less critical thinking and weaker neural connectivity during AI-assisted work. It can also foster overreliance, as users sometimes accept incorrect AI outputs despite conflicting evidence or their own judgement.
Think about what this means. Just as humanity developed machines capable of taking on complex cognitive tasks, humans themselves have become less capable of performing them. The Global Brain Capital Index, which measures cognitive health and related skills, suggests that brain capital is declining at a time when trillions of dollars are being spent on AI infrastructure without a clear idea of what those investments should achieve. And while a recent McKinsey Health Institute analysis estimates that the United States spends roughly $2.6 trillion a year on brain capital, from education to neurological care, less than 1 per cent of that spending goes toward disease prevention. Early intervention and research are underfunded, too.
Technological progress by itself does not make people healthier or more capable. However the AI boom plays out, humanity will not fare well without a global effort to rebuild brain health and cognitive skills. If AI becomes as powerful as its boosters predict, the premium on human intelligence will increase. As AI agents take over routine cognitive tasks, the value of human work will depend on qualities that machines cannot replicate, such as judgement, ethical reasoning, empathy and curiosity.
As the World Economic Forum’s 2025 employer survey showed, many of these human qualities are among those most in demand today and are expected to remain so through 2030. The survey also suggests that 59 per cent of workers will require retraining, making investment in natural intelligence essential if human workers are to remain valuable in an AI-powered economy.
The current debate is largely focused on whether AI will save humanity or destroy it. But it could also turn out to be another technological fad. Even if the AI hype fizzles out like the Metaverse craze of 2022, social and economic progress will still depend on rebuilding brain capital.
We do not need to start from scratch. Much of the necessary infrastructure exists, thanks to those trillions of dollars already being spent on it. The challenge is to treat such investments as part of a larger system. Scaling interventions available today could give people around the world a combined 267 million additional years of healthy life by 2050 and add up to $6.2 trillion to global GDP. Quality early-childhood programs can generate annual returns of 7–13 per cent, while proactive investment in employee health could boost global GDP by as much as 12 per cent.
That means investing in maternal and early-childhood development and shifting primary care toward preventing neurological and psychiatric disorders later in life. Workplaces can protect mental health without eroding cognitive skills, while education and professional training can be overhauled to promote lifelong learning. These are not moonshots: there is plenty of evidence that they work.
AI itself can be part of the solution. It can enable early diagnosis of brain disorders and accelerate drug discovery. By democratising knowledge, it can also feed curiosity and make lifelong learning easier.
The returns on brain capital do not hinge on the fate of the AI industry. Human ingenuity has economic value no matter which technology dominates the next decade. Investing in it is always a winning bet, because it is the one technology that will never become obsolete.
Copyright Project Syndicate 2026
Harris Eyre
Harris Eyre is Executive Director of the Global Brain Economy Initiative.
Lenny Mendonca
Lenny Mendonca, former chief economic and business adviser to Governor Gavin Newsom of California, is Senior Partner Emeritus at McKinsey & Company.
Remington Tonar
Remington Tonar is Co-founder of Cart.com.