Is AI Making Us Less Intelligent? How Chess Can Protect Your Brain in the Age of LLMs

By Amila Rajapakse ( amilaraja@gmail.com)

TL;DR
AI isn't killing IQ, but it's changing how we think through cognitive offloading—outsourcing memory, reasoning, and even social skills to machines. This can weaken critical thinking, attention, and empathy over time. Chess (and similar effortful activities) actively trains these skills and may protect long-term brain health.

Key Takeaways

• AI offloads thinking: Boosts short-term efficiency but risks eroding reasoning, memory, and independent judgment with heavy use.
• Social skills suffer too: Algorithmic feeds + AI chatbots reduce practice reading emotions, navigating disagreement, building real connections.
• Chess = mental gym: Proven to strengthen executive function, attention, working memory; builds cognitive reserve against decline.
• Smart AI use: Think first, AI second. Use as coach/critic, not crutch. Pair with chess, face-to-face talks, analog problem-solving.

Artificial intelligence tools are now woven into everyday work and learning. From writing emails to debugging code, many people rely on large‑language models (LLMs) and other AI assistants to offload thinking tasks. While this brings efficiency, a growing body of research suggests that heavy dependence on AI and other digital tools can weaken specific cognitive skills such as critical thinking, memory, and deep attention. At the same time, evidence indicates that cognitively demanding activities like chess can strengthen executive function, attention, and even long‑term brain health.

This article reviews what we know about AI‑driven cognitive offloading, outlines its potential social consequences, and explains how regular chess practice can serve as a “mental gym” to keep your brain and our societies cognitively resilient.

AI and the erosion of independent thinking

AI tools do not yet appear to be lowering overall IQ scores, but they are changing how people think. A central concept here is cognitive offloading: the tendency to delegate memory, problem solving, and decision‑making to external systems such as search engines, calculators, GPS, and now LLMs.

A 2025 study, AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking, reported a strong positive correlation between AI tool usage and cognitive offloading, and a strong negative correlation between cognitive offloading and critical‑thinking scores (Gerlich, 2025). In practical terms, people who used AI more frequently tended to offload more mental work, and those who offloaded more showed weaker performance on critical‑thinking assessments.

This pattern echoes broader digital‑technology research. A narrative review on digital media and cognition found that constant access to online tools can encourage “shallower and more distracted thinking” and reduce the depth of learning and reflection, especially among heavy users (Shanmugasundaram & Tamilarasu, 2023; de Barros, 2024). Studies on internet search, for example, show that people increasingly remember where to find information rather than the information itself, shifting memory strategies toward external “transactive memory” systems.

The key risk is not that AI immediately makes people less intelligent, but that habitual overreliance on AI encourages passivity and erodes the practice of sustained reasoning, evaluation, and self‑generated problem‑solving.

Cognitive offloading: trends and effects on skills

What the research says about cognitive offloading trends

Recent work synthesizing the impact of digital technology on cognition highlights several consistent trends:

• A 2024 narrative review in IBRO Neuroscience Reports concludes that smartphones, social media, internet search, video gaming, and AI systems all support various forms of cognitive offloading, with mixed consequences for attention, memory, and decision‑making (de Barros, 2024).
• A 2023 review in Frontiers in Cognition notes that constant digital distractions can undermine deep concentration and that multitasking across devices is linked to poorer comprehension and academic performance (Shanmugasundaram & Tamilarasu, 2023).​
• The IEEE Computer Society’s 2025 analysis on cognitive offloading argues that the unprecedented convenience of AI tools creates a “hidden yet profound threat” to human critical thinking by normalizing externalization of evaluation and judgment (Nosta, 2025).

At the same time, a 2025 meta‑analysis on technology use and cognitive aging in Nature Human Behaviour found that everyday digital‑technology use is associated with reduced odds of cognitive impairment and slower cognitive decline in older adults, suggesting that technology can also function as a compensatory tool when used actively and strategically (Benge et al., 2025).

Effects on intelligence‑related skills

Cognitive offloading affects skills more than innate intelligence:

• Critical thinking and reasoning: The AI‑critical‑thinking study mentioned above showed that cognitive offloading mediates the relationship between AI use and lower critical‑thinking scores (Gerlich, 2025), implying that the mechanism is behavioral and potentially reversible.
• Attention and working memory: Smartphone “brain drain” experiments show that the mere presence of one’s phone can reduce available working memory and fluid intelligence, even without active use (Ward et al., 2017), illustrating how digital companions can silently tax cognitive resources.
• Spatial memory and navigation: Long‑term GPS use has been linked to declines in hippocampal‑dependent spatial memory and navigation skills, indicating that external navigation tools can weaken internal spatial representations (Bohbot et al., 2018; Iaria et al., 2020).

Overall, the literature suggests a trade‑off: offloading can free cognitive resources and support performance in complex environments, but if overused for tasks that should remain “core” (reasoning, memory, navigation, deep reading), it can degrade those very skills over time.

Social consequences of widespread cognitive offloading

When cognitive offloading scales from individuals to entire populations, the consequences extend beyond personal performance into social and cultural domains.

1. Vulnerability to misinformation and manipulation

Reduced critical‑thinking capacity at scale can weaken societies’ ability to resist misinformation. The IEEE analysis on cognitive offloading warns that as people grow accustomed to accepting algorithmic outputs uncritically, they become less likely to independently verify information and more vulnerable to persuasive but misleading content (Nosta, 2025).​

A 2023 review on digital technology, AI, and cognition notes that the combination of personalized feeds, filter bubbles, and decreased deep reading can foster narrow, biased worldviews, making it harder for citizens to evaluate political claims and scientific evidence objectively (de Barros, 2024). This raises concerns for democratic decision‑making, where informed, skeptical, and reflective citizens are essential.

2. Shifts in education and skill development

In educational settings, heavy reliance on AI‑driven tools can change what and how students learn. While AI‑based tutoring systems and adaptive platforms can personalize instruction and support struggling learners, they can also encourage “cognitive laziness” if students treat AI as a shortcut rather than a scaffold (Shanmugasundaram & Tamilarasu, 2023).

Reviews of digital technology in education show that multitasking and constant device use are linked to lower grades, reduced comprehension, and superficial learning (Junco & Cotten, 2012; Rosen et al., 2013; summarized in de Barros, 2024). If generative AI becomes the default solution for writing, problem sets, and research, there is a risk that students will graduate with weaker independent writing, reasoning, and information‑literacy skills, even if they remain highly adept at prompting systems.

3. Workplace dependency and loss of adaptive capacity

In the workplace, AI‑driven decision‑support tools can enhance productivity and reduce routine cognitive burdens, but they may also cultivate algorithmic dependency. The IEEE article on cognitive offloading notes that professionals who rely heavily on AI algorithms may struggle when required to make decisions in novel situations where AI tools are unavailable or produce unreliable outputs.​

This dependency can be especially dangerous in high‑stakes fields such as healthcare, law, finance, and critical infrastructure, where overtrust in AI systems could lead to poor decisions, ethical blind spots, or catastrophic failures if systems are biased, attacked, or disrupted (de Barros, 2024). Over time, a workforce that under‑practices manual analysis and domain reasoning may lose adaptive expertise—the ability to improvise and innovate in unfamiliar conditions.

4. Social cognition, empathy, and relationships

Digital‑technology reviews also highlight changes in social cognition. Excessive social‑media use has been associated with decreased grey‑matter volume in emotion‑related brain regions, difficulties in recognizing facial emotions, and a shift toward online interactions at the expense of offline social skills (Montag et al., 2017; Zhou et al., 2011; summarized in de Barros, 2024).​

As AI chatbots and social agents become more capable and ubiquitous, there is concern that some individuals may preferentially interact with AI “companions,” potentially dampening real‑world empathy, conflict‑resolution skills, and nuanced social understanding (Adeyemi, 2025). These trends, combined with cognitive offloading, could yield communities that are simultaneously hyper‑connected and socially fragile.

5. Inequality and cognitive stratification

Access to AI and digital tools is uneven across socioeconomic groups, and so is the capacity to use them critically. Some researchers warn of emerging cognitive stratification, where those with strong foundational skills and good education use AI as an amplifier of their abilities, while those with weaker foundations may become increasingly dependent on AI without developing internal competencies (Benge et al., 2025; Shanmugasundaram & Tamilarasu, 2023).

This can widen gaps in employability, income, and social influence: one group uses AI as a leverage tool; the other uses it as a crutch. In this context, deliberate cognitive training—through activities like chess—can be a way to maintain and democratize core mental skills in an AI‑saturated environment.

Why chess is a powerful cognitive workout

Chess is uniquely suited to counterbalance the kinds of thinking that AI tends to replace: planning, evaluating alternatives, holding complex structures in working memory, and resisting impulsive decisions. A meta‑analysis of chess instruction in schools found small but significant improvements in mathematics achievement and general cognitive ability, suggesting that chess can transfer benefits to academic domains (Sala & Gobet, 2017).​

Graph‑theory work on adult chess players shows that they exhibit more integrated “cognitive connectomes,” with central roles for nodes related to executive function, attention, processing speed, and visuospatial abilities (Fattahi et al., 2024). Intervention studies with children indicate that an 8‑week blended chess program can improve selective and sustained attention and certain academic skills (Bart, 2023).

In short, chess trains:

• Working memory (holding multiple candidate moves and lines).
• Executive control (inhibiting impulsive moves, sticking to a plan).
• Cognitive flexibility (switching plans rapidly as positions change).
• Pattern recognition (identifying recurring structures and motifs).

These are exactly the capacities that risk degradation under habitual cognitive offloading and that societies need for robust democratic participation, innovation, and problem solving.

Chess, cognitive reserve, and long‑term brain health

Beyond immediate cognitive gains, chess may also support long‑term brain health. A scoping review on chess and dementia concluded that regular chess practice can be considered a protective factor against cognitive decline in non‑diagnosed populations, likely by enhancing cognitive reserve (Lillo‑Crespo et al., 2019).​

The review and related cohort studies show that:

• Older adults who frequently engage in board games are significantly less likely to develop dementia than those who rarely do.​
• Participation in multiple mind‑exercising activities, including chess, is associated with delayed onset of Alzheimer’s‑related symptoms by about 1.5 years (Barnes et al., 2013; summarized in Lillo‑Crespo et al., 2019).​

A 2023 article from Harvard highlights ongoing longitudinal work using chess to probe how sustained strategic gameplay may protect brain health, especially in aging populations (Harvard Gazette, 2023). Taken together, these findings support the idea that chess can help maintain both individual cognitive resilience and, at scale, the collective cognitive capacity societies will need in an AI‑heavy future.

A practical “AI‑age chess protocol” for brain and society

If AI offloading reduces the mental load on your brain, you can treat serious chess practice as deliberate strength training for the cognitive muscles you and your community might otherwise stop using.

1. Personal practice

• Play chess 3–5 times per week for 30–60 minutes with longer time controls to encourage deep thinking rather than reflex moves countering the fragmented attention AI habits create rather than adding to digital chaos.
• Mix slow games, tactical puzzles, and post‑game self‑analysis before using engines to foster metacognition and resist passive pattern‑copying.

2. Educational and community programs

• Integrate chess into school curricula as a tool for teaching problem solving, planning, and emotional regulation, not just as a club activity.
• Use chess clubs and community tournaments to create spaces where children and adults practice focused, device‑free strategic interaction, countering fragmented attention and shallow online engagement.

3. Mindful AI usage norms

• Encourage a “human‑first, AI‑second” workflow: attempt a solution yourself, then use AI to critique, expand, or challenge your reasoning.
• In professional training, pair AI‑based tools with exercises that require “manual” analysis and justification to maintain domain expertise and adaptive capacity.

By combining intentional AI use with regular chess practice and other cognitively demanding, socially grounded activities, individuals and communities can protect both personal intelligence‑related skills and the social fabric that depends on them.

References

• Adeyemi, V. (2025). The impact of digital technology on a child’s cognitive and social development: A systematic review. Philippine Journal of Psychology, 25(2), 808–832.​
• Benge, J. F., Scullin, M. K., & colleagues. (2025). A meta‑analysis of technology use and cognitive aging. Nature Human Behaviour, 9(4), 612–628.
• de Barros, E. C. (2024). Understanding the influence of digital technology on human cognitive functions: A narrative review. IBRO Neuroscience Reports, 16, 100201.
• Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. Journal of Societal Contexts and Technology, 15(1), 6.
• Nosta, J. (2025). Cognitive offloading: How AI is quietly eroding our critical thinking. IEEE Computer Society Tech News.
• Shanmugasundaram, M., & Tamilarasu, K. (2023). The impact of digital technology, social media, and artificial intelligence on human cognition. Frontiers in Cognition, 2, 1203077.
• ​Sala, G., et al. (2017). The Effects of Chess Instruction on Pupils' Cognitive and Academic Skills. PMC.​
• Frontiers in Psychology (2024). The effect of chess on cognition: a graph theory study.​
• De Groot et al. (2019). Chess Practice as a Protective Factor in Dementia. PMC.​