We have never had so much information at our fingertips, yet there are growing reasons to wonder whether we are becoming less capable of making sense of it. A smartphone can retrieve centuries of human knowledge in seconds. Artificial intelligence can explain complex theories, construct arguments and write convincing essays. Meanwhile, millions of people struggle to distinguish reliable information from manipulation, public debate increasingly resembles a competition in outrage, and confidence often seems to count for more than competence. The uncomfortable possibility is that our technological progress has outpaced our ability, or perhaps our willingness, to think. There is evidence that some of the skills associated with critical thinking are weakening.
In December 2024, the Organisation for Economic Co-operation and Development published findings from its Survey of Adult Skills, covering approximately 160,000 people across 31 countries and economies. Literacy and numeracy had declined or stagnated across much of the developed world over the preceding decade. Only Finland and Denmark recorded significant improvements in adult literacy among the countries compared. More education, better technology and easier access to information had not necessarily produced stronger fundamental skills.
The picture becomes more troubling when we examine younger generations. American educational assessments revealed declining reading performance before the pandemic, followed by further deterioration. In 2024, the average reading scores of fourth- and eighth-grade pupils fell again. By 2025, just 14% of 13-year-olds reported reading for pleasure almost every day, compared with 27% in 2012. Yet the evidence is not uniformly negative. American nine-year-olds improved their reading and mathematics results between 2022 and 2025, while England was among the few countries where the OECD identified improvements in young adults’ literacy. We are observing worrying trends, not a universal intellectual collapse.
Nor should we confuse literacy with critical thinking. Someone may read fluently and still believe an absurd conspiracy theory. Another person might struggle with formal education yet possess excellent practical judgement. Critical thinking involves questioning assumptions, evaluating evidence, recognising contradictions and changing one’s mind when the facts demand it. These abilities are difficult to measure across entire societies. The available statistics tell us something about the foundations of reasoning, but they cannot establish that humanity has collectively become less intelligent.
An especially revealing experiment came from researchers at Stanford University, who assessed the online reasoning skills of 3,446 American high-school students. Participants could use the internet to investigate information presented to them. When examining a website that claimed to provide factual reporting about climate science, 96% failed to identify its connections to the fossil fuel industry. The problem was not a shortage of available information. The students had access to the very tool that could have exposed the connection. What they lacked was an effective method for questioning the source.
This illustrates an overlooked problem with the information age. Access to knowledge and the ability to evaluate knowledge are entirely different achievements. Search engines removed much of the effort involved in finding an answer, but they did not remove the need to decide whether that answer deserves belief. In some respects, they made the second task harder. An unreliable claim can appear alongside legitimate academic research, presented with comparable confidence and professional design. The internet has democratised publication without democratising expertise, leaving readers to perform an editorial function they may never have learnt.
Social media adds another complication. Its commercial incentives frequently reward attention rather than accuracy. A provocative claim can generate more engagement than a careful explanation, especially when it confirms an existing belief. Researchers Gordon Pennycook and David Rand examined 20 experiments involving more than 26,000 participants and found that simple prompts encouraging people to consider accuracy could reduce their willingness to share misinformation. This suggests that at least part of the problem is not an inability to recognise falsehoods. People sometimes neglect to consider whether information is true because their attention has shifted towards something else.
That distinction matters. Someone sharing an outrageous headline may be expressing loyalty to a group, demonstrating moral outrage or entertaining friends rather than deliberately assessing facts. The social function of a statement can overwhelm its informational value. In political discussions, this becomes especially damaging. Disagreement stops being an opportunity to examine competing explanations and becomes a test of allegiance. Admitting uncertainty may sound weak; changing one’s mind can look like betrayal. Certainty has become a social performance, and occasionally a remarkably successful career strategy.
Artificial intelligence introduces another layer. In 2025, researchers from Microsoft and Carnegie Mellon University surveyed 319 knowledge workers about their experiences using generative AI. They collected 936 examples of workplace tasks and found an association between greater confidence in AI and less reported critical thinking. Workers who trusted their own abilities tended to report more critical engagement. Importantly, the study examined self-reported behaviour rather than proving that AI causes lasting cognitive decline. It also found that AI could shift intellectual effort towards checking information and integrating results.
The potential danger is nevertheless apparent. When a machine produces a polished argument in seconds, the temptation is to assess its presentation rather than reconstruct its reasoning. A confident explanation can conceal a questionable assumption, an invented reference or a logical gap. A beautifully structured business proposal may become particularly persuasive because it looks finished. Yet the hardest intellectual work often happens before the document exists: deciding what problem actually needs solving, which assumptions deserve testing and what evidence would invalidate the proposed solution.
There is a broader historical irony here. Socrates, as portrayed in Plato’s Phaedrus, expressed concerns that writing would encourage people to rely on external records rather than exercise their memories. Similar anxieties accompanied printing, calculators and computers. Some proved exaggerated, while others identified genuine changes in how people use their minds. Technologies can relieve us of routine mental labour and free our attention for harder questions. The critical issue is whether we use that freedom to think more deeply or simply to avoid thinking altogether.
Education also deserves scrutiny. Schools and universities often reward correct answers, examination performance and the ability to reproduce established material. Those achievements have value, but they do not automatically teach intellectual independence. A student who learns to defend an argument against serious objections may develop better reasoning habits than one who memorises twenty perfectly formulated conclusions. Equally, teaching people to question everything without giving them the knowledge needed to judge competing claims can produce cynicism rather than discernment.
Perhaps the most revealing measure of critical thinking is not how quickly someone answers a difficult question but how they respond when confronted with evidence that contradicts their convictions. Do they investigate it, dismiss it or search for reassurance from people who already agree with them? Genuine scepticism applies the same standards to agreeable and disagreeable claims. It involves resisting both blind trust in authority and the equally unhelpful assumption that every expert must be wrong. Neither obedience nor automatic contrarianism requires much reasoning.
The encouraging implication is that critical thinking can be practised through ordinary habits. Reading complete arguments rather than headlines, comparing independent accounts, examining original evidence and deliberately considering alternative explanations all create opportunities for better judgement. So does the simple discipline of distinguishing what we know from what we assume. These practices are neither glamorous nor particularly compatible with a permanent sense of urgency. They require time, attention and a willingness to remain uncertain.
We may therefore be facing something subtler than a decline in intelligence. Modern society has created an environment in which independent reasoning is increasingly easy to bypass. Information arrives already selected, arguments already packaged and answers already written. The danger is not that machines will become capable of thinking. It is that human beings may gradually lose the habit of asking whether the thinking has been done properly.
