The Dual-Edged Sword of AI Tools: Balancing Cognitive Offloading and Critical Thinking in an AI-Driven Era

Soundcloud Podcast

Executive Summary

This article explores the intricate relationship between artificial intelligence (AI) tools and human cognition, focusing on cognitive offloading and critical thinking. As AI becomes an integral part of daily life, its impact on mental processes like decision-making, memory, and problem-solving warrants vital examination. While AI offers efficiency and innovation, it also poses risks of dependency and cognitive skill erosion, particularly among younger generations.

Key Findings

1. The Dual Nature of AI Tools:
   AI tools, particularly those powered by large language models and generative systems, enhance efficiency in repetitive and computational tasks. However, over-reliance on AI fosters cognitive offloading, which reduces engagement in deeper reflective thinking and critical analysis.
2. Role of Neuromodulators:
   Neuromodulators like dopamine, acetylcholine, and norepinephrine are pivotal in determining whether individuals engage with tasks or offload them to AI. Supplements targeting these neuromodulators can influence cognitive engagement but require careful use to avoid fostering dependency.
3. Critical Thinking and Education:
   Depending on their use, AI tools can erode or enhance critical thinking. Tasks such as analyzing and critiquing AI-generated content encourage cognitive resilience. Educational strategies must prioritize teaching students to engage critically with AI outputs, ensuring that human oversight and judgment remain central.
4. Ethical and Design Considerations:
   AI system developers must design tools that promote active engagement and transparency, offering features that contextualize and explain AI outputs. Policymakers should create guidelines that ensure equitable access to AI technologies while preserving human oversight in decision-making processes.
5. Future Research:
   Longitudinal studies are needed to explore the cumulative effects of AI reliance on cognition over time. Investigating demographic factors such as age, socioeconomic status, and education level can provide insights into how different populations are affected.

Recommendations

1. For Educators:
   Integrate assignments that encourage students to evaluate, refine, and improve upon AI-generated content. Teach AI literacy alongside critical thinking skills to prepare students for an AI-driven world.
2. For Developers:
   Design AI tools that foster user engagement and critical evaluation rather than passive acceptance. Incorporate features that prompt users to trace the reasoning behind AI outputs and verify their accuracy.
3. For Policymakers:
   Establish ethical frameworks that prioritize transparency and equitable AI access. Promote interdisciplinary collaboration to ensure that AI systems complement, rather than replace, human cognition.

Conclusion

AI’s integration into society is irreversible, but its impact on human cognition remains a choice. By balancing leveraging AI’s strengths and preserving cognitive independence, society can ensure that AI becomes an ally in human progress. The key lies in fostering a deliberate, thoughtful engagement with AI tools that empower individuals to think critically, act creatively, and retain agency in the face of automation.

The future of cognition depends not only on the capabilities of AI but also on our collective choices in using it. This balance between reliance and engagement will define whether AI becomes a catalyst for human growth or a harbinger of cognitive decline.

Introduction

The rise of artificial intelligence (AI) tools marks a pivotal turning point in human history, reshaping how we think, create, and interact. From virtual assistants managing daily tasks to generative systems producing intricate works of art and writing, AI has woven itself into the fabric of everyday life. With advanced technologies like transformer-based architectures and large language models (LLMs), AI can process, generate, and synthesize information at speeds that challenge human imagination.

Yet, alongside its transformative potential comes a pressing need for scrutiny. These tools promise unprecedented efficiency and innovation, enabling us to offload repetitive or computational tasks to machines. But this act of cognitive offloading — the delegation of mental functions to external aids — raises profound questions about its impact on critical thinking, problem-solving, and memory. What happens when reliance on AI extends beyond convenience and begins to erode the core cognitive processes that make us human?

This dual-edged nature of AI integration demands a careful balancing act. On one side lies the potential for AI to amplify human cognition, fostering creativity, personalization, and new opportunities for innovation. On the other lies the risk of dependency, cognitive laziness, and the loss of essential skills. Central to this tension are biological mechanisms like neuromodulators, which influence our capacity for engagement, focus, and adaptability. These unseen forces are critical in shaping how we interact with AI and determine when to offload or engage.

The challenge is not merely technological but societal. It requires rethinking education to cultivate critical engagement with AI, designing systems that foster collaboration rather than passive consumption, and establishing ethical frameworks to guide AI’s role in daily life. This article explores these themes, analyzing how AI tools affect human cognition, the role of neuromodulators in decision-making, and the strategies needed to ensure AI empowers rather than diminishes the human mind. At stake is not just the future of technology but the essence of thinking, learning, and creating.

AI Tools and Their Ubiquity: Beyond Everyday Automation

Artificial Intelligence (AI) has become deeply embedded in modern life, often operating seamlessly in the background of our daily routines. Traditional AI systems, such as navigation apps that predict traffic with precision, autocorrect tools that refine our messages, and recommendation algorithms that shape our consumption habits, exemplify the integration of automation into our lives. These tools improve efficiency by addressing routine tasks, enabling users to allocate their attention elsewhere.

60% of professionals reported weekly use of AI tools like ChatGPT in 2024

However, the current wave of AI is defined by transformative advancements in machine learning, particularly transformer-based architectures and large language models (LLMs) such as GPT-4 and similar generative systems. These models employ multi-headed attention mechanisms to parse context, discern patterns, and generate human-like text, images, and code. By leveraging extensive training on massive datasets, generative AI tools can perform tasks ranging from drafting complex documents to providing personalized learning experiences.

These capabilities extend beyond simple automation, offering profound enhancements to creativity and productivity. For instance, over 60% of professionals reported weekly use of AI tools like ChatGPT in 2024, citing their utility in brainstorming, summarizing, and problem-solving tasks (Hinduja, 2024). Similarly, generative AI platforms have become indispensable for functions that demand rapid ideation or detailed synthesis of information.

Yet this sophistication brings new challenges. The rise of cognitive offloading — delegating mental tasks to external aids — has profound implications for human cognition. Historically limited to more straightforward tools like calculators or handwritten notes, cognitive offloading has now expanded to encompass tasks like writing, decision-making, and complex reasoning, traditionally considered the pinnacle of human cognitive engagement. This raises a pressing question: does delegating such tasks to AI systems weaken our capacity for critical thought, or does it liberate our minds to focus on higher-order reasoning and creativity?

Cognitive Offloading: A Double-Edged Phenomenon

Cognitive offloading—the act of delegating mental tasks to external aids—is as old as human innovation itself. The invention of writing allowed memories to be etched onto stone tablets, and the calculator unshackled us from the tedium of arithmetic. These tools were hailed as triumphs of human ingenuity, freeing mental resources for higher pursuits. Yet, what began as a means to transcend cognitive limits has evolved into a phenomenon that reshapes the very nature of thought.

At its core, cognitive offloading is not inherently harmful. It is, in fact, the foundation of progress. By relying on external tools to handle repetitive or menial tasks, individuals can channel their cognitive resources into creativity, problem-solving, and analytical reasoning (Risko & Dunn, 2015). The efficiency gains are undeniable. A student who uses a calculator can solve equations faster and with fewer errors, leaving time for theoretical exploration. A writer using AI-powered suggestions can draft faster, allowing more focus on the coherence of ideas.

However, the same tools that promise liberation also pose subtle dangers. The human brain, unlike machines, is shaped by use. When it encounters a task repeatedly, neural pathways strengthen; skills are sharpened. Offloading certain types of cognitive tasks diminishes this reinforcement. The “Google effect” phenomenon identified by Sparrow, Liu, and Wegner (2011) illustrates this well. In their research, participants remembered the locations where information was stored rather than the information itself, relying on the expectation of future accessibility. This shift reflects a broader trend: as we lean more toward external systems, we engage less deeply with the content of our thoughts.

What the Brain Can — and Cannot — Offload

The human mind is adept at offloading formulaic, repetitive tasks or requiring high precision. Navigation, for example, is easily surrendered to a GPS device, which calculates the fastest route in milliseconds. Similarly, rote memorization can be delegated to notes, search engines, or AI-powered flashcard systems. These tools excel because they handle predictable tasks faster and more accurately than a distracted brain ever could.

But the brain cannot outsource everything. It cannot hand over judgment, creativity, or the ability to make sense of nuance. A navigation app may provide directions, but deciding whether to detour for a scenic view is a uniquely human choice. Similarly, an AI tool may generate coherent text, but recognizing when that text lacks ethical integrity or cultural sensitivity requires human oversight. Judgment is not formulaic but an emergent property of context, experience, and reflection.

10–15% decline in deductive reasoning, inference-making, and similar higher-order skills over the past three decades

The inability to offload judgment or creativity underscores the irreplaceability of critical thinking. These faculties depend on knowledge and how that knowledge is organized, interpreted, and applied. When we rely too heavily on AI for these processes, we risk weakening the skills that define them. Recent research highlights this danger. A meta-analysis spanning 60 studies revealed a 10–15% decline in deductive reasoning, inference-making, and similar higher-order skills over the past three decades — a trend correlated with growing technological reliance (Risko & Gilbert, 2023).

The Long Shadow of Cognitive Offloading

The consequences of over-reliance on cognitive offloading extend beyond individual cognition to society. Younger generations, raised in a world saturated with AI tools, show the most pronounced effects. Studies suggest that these individuals exhibit higher reliance rates on AI and score lower on critical thinking assessments (Gerlich, 2025). This correlation hints at a troubling feedback loop: critical thinking weakens as reliance on AI increases, leading to even greater dependence on external systems.

This erosion of skills is not easily reversed. Neural pathways that atrophy from disuse take time and deliberate effort to rebuild. Consider the example of memorization. Historically, scholars honed their minds through rote learning, committing vast amounts of poetry, scripture, or historical dates to memory. Today, such practices are often dismissed as outdated, with search engines providing instant access to any fact. Yet studies show memorization strengthens cognitive resilience, forming a mental framework that facilitates deeper analysis and synthesis (Heersmink, 2020).

Striking a Balance

The solution is not to abandon cognitive offloading altogether but to wield it with intentionality and discernment. Cognitive offloading is an inevitable aspect of modern life; its benefits are undeniable in scenarios that require precision, speed, or vast computational capabilities. However, the real question is not whether to offload tasks to AI but how and when to do so in a way that preserves and enhances human cognition.

Balancing Machine Efficiency with Human Judgment

Tasks that involve repetitive computation, data retrieval, or systematic categorization are undoubtedly better suited to machines. Navigation apps that calculate the fastest route or AI tools that generate comprehensive text summaries exemplify offloading, which improves efficiency without diminishing human creativity. Yet, even in these instances, the risks of dependency loom large. What happens when GPS navigation removes our spatial awareness or reliance on AI summaries diminishes our ability to discern subtleties in information?

The answer lies in calibrated reliance: reserving offloading for tasks that do not inherently contribute to developing critical thinking or judgment. For instance, using AI to organize information is a productive form of offloading, but analyzing that information and applying it to nuanced contexts must remain firmly within the domain of human cognition.

Neuromodulators: The Chemical Governors of Cognitive Engagement

At the heart of this balance lies the brain’s internal machinery, particularly the neuromodulators that govern motivation, attention, and flexibility. Dopamine, acetylcholine, and norepinephrine play a silent but pivotal role in determining whether we engage deeply with a task or delegate it to an external aid.

1. Dopamine: The Effort-Reward ArbiterHigh dopamine levels encourage engagement with cognitively demanding tasks, while low levels increase the allure of offloading to reduce mental strain. Individuals with depleted dopamine reserves tend to avoid challenging problems, turning to AI tools for easy answers.
2. Acetylcholine: The Focus FacilitatorAcetylcholine amplifies selective attention, enabling deep engagement with information. When acetylcholine levels are high, tasks requiring detailed analysis and learning become more manageable, reducing the need for offloading. Conversely, a lack of acetylcholine fosters superficial engagement, increasing reliance on external aids.
3. Norepinephrine: The Adaptability EnablerNorepinephrine regulates alertness and adaptability. Balanced norepinephrine levels support thoughtful decisions about when to engage with a task and when offloading is appropriate. Dysregulated levels—either too high or too low—impair this adaptability, leading to an overreliance on external tools or a refusal to offload when it would be beneficial.

The Promise and Perils of Supplements

Modernity’s answer to faltering neuromodulation comes in the form of supplements. These chemical aids target the brain’s natural systems, offering the promise of sharper focus, greater motivation, and enhanced adaptability. However, their efficacy is not without limits, and their use must be approached cautiously.

1. Dopamine-Enhancing Supplements
   L-Tyrosine, a dopamine precursor, has shown promise in boosting motivation during periods of stress. Similarly, Rhodiola Rosea reduces mental fatigue, encouraging engagement rather than offloading. Yet reliance on such supplements raises questions: does chemically induced motivation lead to sustainable engagement, or does it foster dependency?
2. Acetylcholine-Boosting Supplements
   Citicoline and Huperzine A enhance acetylcholine levels, promoting attention and learning. These compounds reduce the need to offload by enabling deeper cognitive engagement. However, their benefits are temporary and cannot replace the foundational critical thinking habits.
3. Norepinephrine-Regulating Supplements
   Ginseng and combining L-Theanine and Caffeine support norepinephrine activity, aiding decision-making and cognitive flexibility. These supplements may improve the ability to discern when offloading is appropriate, but like all interventions, they must be used as tools, not crutches.

While supplements may enhance neuromodulator activity, they cannot replace the long-term benefits of cognitive discipline, mindfulness, and well-designed educational strategies.

Educational Strategies for Critical Engagement

Education must serve as the crucible in which students learn how to use AI tools and think critically about their outputs. While often dismissed as outdated, traditional rote learning methods had the unintended benefit of strengthening neural pathways critical for memory and problem-solving. Modern educators face the challenge of cultivating these same pathways in an era dominated by instant access to information.

Assignments that emphasize critical engagement with AI-generated content can address this challenge. For example:

• Students might be tasked with identifying errors, biases, or gaps in AI-produced summaries.
• Interactive discussions involve debating the ethical implications of AI-generated outputs, requiring students to synthesize information and argue their perspectives.
• Group projects could focus on comparing human and AI approaches to problem-solving, fostering an understanding of these tools’ capabilities and limitations.

These activities shift AI’s role from that of an unquestioned authority to that of a collaborative partner. By actively engaging with AI outputs, students retain agency in learning and develop the cognitive resilience necessary for critical thinking.

The Role of AI Developers in Preserving Human Cognition

The responsibility for striking this balance does not rest solely with educators. Developers of AI systems bear a significant share of the burden. Too often, AI tools are designed to minimize user effort, creating an illusion of simplicity that discourages critical engagement. To foster a more symbiotic relationship between humans and machines, developers must prioritize features that:

• Encourage Contextualization: Tools should allow users to trace the sources and reasoning behind AI-generated outputs. For example, language models could display their reasoning pathways or offer multiple interpretations of ambiguous prompts.
• Facilitate Active Input: Systems should reward user engagement by offering more sophisticated outputs in response to thoughtful queries.
• Prompt Critical Reflection: Developers could incorporate features that explicitly ask users to verify or critique outputs.

The Risk of Miscalibration

The consequences of failing to strike this balance are dire. When over-reliance on AI leads to cognitive laziness, the erosion of critical thinking becomes self-reinforcing. A workforce that depends on AI for judgment and creativity risks becoming passive and unable to innovate or adapt when technology fails. Societies that abdicate decision-making to algorithms face a similar fate: decisions shaped by unseen biases and blind adherence to machine logic rather than informed human reasoning.

Conversely, avoiding cognitive offloading altogether is equally untenable. Rejecting AI’s capabilities to preserve human cognition would squander opportunities for efficiency, innovation, and progress. The key is to offload intelligently, leveraging AI to complement human strengths rather than supplant them.

A Framework for Balance

Achieving this balance requires a framework that integrates education, policy, and technology design:

• In Education: Reform curricula to prioritize skills such as critical evaluation, ethical reasoning, and information synthesis. Teach students to see AI as a tool to be mastered, not a crutch to be leaned upon.
• In Policy: Establish guidelines for ethical AI deployment, emphasizing transparency and the preservation of human oversight in decision-making.
• In Technology: Design AI systems that are transparent, interpretable, and conducive to user engagement. Features encouraging curiosity, scrutiny, and reflection should be prioritized over those promoting passive consumption.

Ultimately, the goal is not to resist cognitive offloading but to wield it wisely. Tasks that demand speed and accuracy can be offloaded, provided the underlying cognitive and ethical processes remain active. We can ensure that AI serves as a catalyst for human growth rather than a substitute for human thought by maintaining this balance. The path to a future where technology and cognition flourish together lies in this delicate dance between reliance and engagement.Conclusion

Cognitive offloading is neither a savior nor a scourge. It is a tool, and like any tool, its value lies in how it is used. Properly managed, it can amplify human potential, freeing us to explore realms of thought previously unreachable. Mismanaged, it risks hollowing out the skills that define us, leaving us dependent on machines we no longer understand. The challenge is not to reject cognitive offloading but to master it — to use it to serve a richer, more reflective human experience. The Potential of AI to Enhance Critical Thinking

Yet, to dismiss AI tools as mere agents of cognitive decline would be shortsighted. When used strategically, AI has the potential to enhance critical thinking rather than diminish it. In education, for instance, adaptive learning platforms like Kialo Edu foster engagement by tailoring content to individual needs. Interactive simulations challenge students to apply abstract concepts to realistic scenarios, while AI-generated feedback encourages iterative improvement (Kaledio et al., 2024).

A study of English language learners found that AI-based tools improved language proficiency and bolstered critical thinking by prompting users to analyze and critique AI-generated outputs (Lawasi, Rohman, & Shoreamanis, 2024). However, the researchers noted a caveat: the effectiveness of these tools depends on the user’s ability to ask precise questions and interpret responses critically. Without these skills, AI becomes a crutch rather than a catalyst for intellectual growth.

The Ethical Quandaries of AI

The promise of AI cannot be fully realized without addressing its ethical implications. Bias in AI algorithms, lack of transparency, and the potential for misuse are well-documented concerns (Rahman & Watanobe, 2023). Moreover, the democratization of AI tools risks widening cognitive disparities. Those with access to high-quality AI systems will benefit from enhanced learning opportunities, while others may fall further behind. This digital divide threatens to exacerbate existing inequities in education and the workforce.

Furthermore, AI’s influence on decision-making raises questions about autonomy and accountability. Automated systems in healthcare and finance have streamlined operations but at the cost of reducing human oversight. Professionals who defer to AI recommendations without scrutiny risk becoming passive participants in their fields (Pasquale, 2020).

Striking a Balance: Strategies for Responsible AI Integration

The challenge is to harness AI’s capabilities without succumbing to its pitfalls. Achieving this balance requires a multifaceted approach.

1. Designing Human-Centered AI: Developers must prioritize transparency and interpretability in AI systems. Tools should encourage user engagement rather than passive consumption, prompting users to verify and contextualize AI-generated outputs (Gerlich, 2025).
2. Educational Interventions: Curricula must evolve to include AI literacy as a core component. Students should learn not only how to use AI tools but also how to evaluate their outputs critically. Assignments that require students to analyze and improve upon AI-generated content can foster deeper cognitive engagement (Hinduja, 2024).
3. Promoting Reflective Practices: Reflective practices should be encouraged in education and the workplace. Activities that require deep analysis, problem-solving, and independent decision-making can counteract the cognitive inertia induced by AI (Lawasi et al., 2024).
4. Policy and Regulation: Policymakers must establish guidelines to ensure equitable access to AI technologies and protect against misuse. Ethical standards for AI deployment in sensitive fields like education and healthcare are essential to maintaining trust and accountability (Rahman & Watanobe, 2023).

Future Directions

The interplay between AI and human cognition remains a fertile ground for research, with much left to uncover. To fully understand AI’s long-term impact on cognitive processes, the following directions warrant immediate attention:

1. Longitudinal Studies on AI Reliance:
   Short-term studies offer glimpses into AI’s effects. Still, only longitudinal research can reveal how decades of reliance on AI tools shape cognitive abilities such as memory retention, problem-solving, and critical thinking. These studies must account for generational shifts and capture how individuals' cognitive profiles evolve in an AI-dominated world.
2. Demographic Influences on AI-Cognition Dynamics:
   Demographics such as age, education level, socioeconomic status, and cultural background may significantly mediate the relationship between AI usage and cognitive resilience. Younger users, raised with ubiquitous technology, might exhibit different patterns of cognitive engagement compared to older generations. Similarly, access disparities driven by socioeconomic status could either exacerbate or mitigate AI’s cognitive impact.
3. The Role of Neuromodulators in AI Engagement:
   Research should explore how AI reliance interacts with brain chemistry, particularly neuromodulators like dopamine and acetylcholine. Understanding these biological mechanisms can provide actionable insights into how cognitive engagement might be preserved or enhanced through behavioral interventions or supplements.
4. Ethical Frameworks for AI-Centric Education:
   With AI increasingly integrated into educational environments, studies should focus on the ethical dimensions of AI deployment. This includes assessing how AI-driven tools influence pedagogy, equity, and the balance between human instruction and automated assistance.
5. Designing Human-Centric AI Systems:
   Collaboration among psychologists, educators, and technologists is crucial for creating AI systems that complement, rather than compromise, human cognition. These systems should prioritize transparency, user engagement, and features that prompt critical thinking. Future research must evaluate how design choices — such as offering contextual explanations for AI outputs — affect cognitive outcomes.
6. Exploring AI’s Role in Creativity and Innovation:
   While much attention is given to AI’s efficiency, its potential to amplify human creativity remains underexplored. Research could investigate how tools like generative AI influence fields that demand high levels of innovation, such as the arts, engineering, and scientific discovery.

Conclusion

The integration of AI into society is an irreversible reality, but its cognitive implications remain ours to shape. AI stands at a crossroads: it can either be an enabler of human potential or a mechanism of dependency and erosion. This dual-edged nature demands vigilance, creativity, and ethical foresight.

To tread this path wisely, we must adopt a balanced approach that leverages AI’s strengths while safeguarding cognitive independence. This involves:

• Educators foster critical thinking alongside AI literacy, ensuring students engage with technology actively, not passively.
• Technologists design systems that invite scrutiny, encourage curiosity, and preserve human oversight.
• Policymakers establish ethical frameworks for AI use, prioritizing transparency and equitable access.

The future of cognition will not be dictated by algorithms alone but by our choices in integrating AI into our lives. Society is on the onus to ensure that this integration empowers rather than diminishes us. Whether AI becomes a tool of empowerment or a harbinger of decline will depend on how we engage with it—deliberately, thoughtfully, and with an unwavering commitment to preserving what makes us human: our capacity to think, create, and question.

In the end, AI's true power lies not in its capacity to replicate human thought but in its ability to enhance and amplify it. Our task is to ensure that we remain the masters of our tools, not their servants. The road ahead is both fraught with challenges and rich with promise. It is ours to navigate with care and conviction.

References

Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. Societies, 15(6). https://doi.org/10.3390/soc15010006

Hinduja, N. (2024). AI innovations to watch in 2024: Transforming everyday life. Dev.to. https://dev.to/namratahinduja/ai-innovations-to-watch-in-2024-transforming-everyday-life-le4

Kaledio, P., Robert, A., & Frank, L. (2024). The impact of artificial intelligence on students’ learning experience. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4716747

Lawasi, M. C., Rohman, V. A., & Shoreamanis, M. (2024). The use of AI in improving students’ critical thinking skills. Proceedings Series on Social Sciences & Humanities, 18. https://doi.org/10.30595/pssh.v18i.1279

McMorris, T., Harris, R. C., & Swain, J. (2020). Tyrosine supplementation and cognitive performance under stress: A review. European Journal of Applied Physiology, 120(5), 1069–1079. https://doi.org/10.1007/s00421-020-04319-6

Pasquale, F. (2020). The black box society: The secret algorithms that control money and information. Harvard University Press.

Risko, E. F., & Dunn, T. L. (2015). Cognitive offloading is value-based decision making: Modeling. Psychological Science, 26(6), 918–926. https://doi.org/10.1177/0956797615574696

Risko, E. F., & Gilbert, S. J. (2023). Cognitive offloading and its implications for understanding cognitive processes. Annual Review of Psychology, 74, 187–209. https://doi.org/10.1146/annurev-psych-052021-093451

Sparrow, B., Liu, J., & Wegner, D. M. (2011). Google effects on memory: Cognitive consequences of having information at our fingertips. Science, 333(6043), 776–778. https://doi.org/10.1126/science.1207745

Westbrook, A., & Braver, T. S. (2015). Dopamine and cognitive control: Beyond motivation and reward. Biological Psychiatry, 78(5), 321–328. https://doi.org/10.1016/j.biopsych.2015.03.014

Yang, T., Wang, H., & He, Y. (2022). The effects of Huperzine A on learning and memory: A systematic review. Neuroscience Letters, 770, 136341. https://doi.org/10.1016/j.neulet.2022.136341

Reay, J. L., Kennedy, D. O., & Scholey, A. B. (2005). Single doses of Panax ginseng (G115) reduce blood glucose levels and improve cognitive performance during sustained mental activity. Journal of Psychopharmacology, 19(4), 357–365. https://doi.org/10.1177/0269881105053286