Based on current fossil evidence, Homo sapiens emerged roughly 300 thousand years ago.1 With early members of our genus emerging around 2 million years ago, and our distant ancestors extending as far back as 7 million years,2,3 humanity represents a mere blip in the 4.6 billion years of Earth's existence.4 Despite this, we have, in short order, become the dominant life form on Earth—a feat attributed to our intellectual capacity.
From cutting food with stone tools to correcting errors in our genetic code, scientific and technological advances have shaped our trajectory as a species. As we continue to develop increasingly advanced tools to springboard into the future, artificial intelligence (AI) is poised to unify the vast corpus of human knowledge. Alongside parallel advances in computational power, AI will continually accelerate development across all facets of our lives. While human health and consumer technology are often the first uses to come to mind, AI already is driving immediate impact across commercial operations.
Unlocking focus and efficiency
In today's dynamic and competitive business environment, efficiency and quality are paramount for success, regardless of the specific sector. To this end, we have replaced our stone hammers with software tools to crush the competition. How do they do this? Mundane, repetitive, and resource-intensive tasks erode profit margins and morale alike. At the same time, there is an opportunity cost in burdening employees with these tasks and taking them away from those tasks that provide greater value to the organization. Invoicing, data entry, timeline generation, and structured but untemplated emails, for instance, are ideal tasks rich with ROI ready to be harvested through automation. While software automation can operate independently of AI, the emergence of the latter has catapulted both the type and complexity of tasks that can be automated.
Smarter search, faster output
The last few years have ushered in an arms race of generative AI Large Language Models (LLMs) and subsequent platforms that support their implementation. Each LLM iteration aims to dig deeper, process faster, and tackle increasingly complex tasks accurately and efficiently. Generative AI-based search tools such as Perplexity and Google-owned Gemini streamline research, while internal systems such as Google NotebookLM allow the interrogation of curated content with a high level of accountability and quality control—provided the user manuallyvalidates the output.
In other words: Tasks that used to take hours or days, such as researching the therapeutic landscape of metastatic non-small cell lung cancer (NSCLC), can now be done in a fraction of the time with a well-written query and a single cup of coffee. Did you also need to dive into its demographics and prevalence? Just tweak the prompt to request that the AI focus on the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) data. Would it help if you had the content distilled into a one-page summary that follows a rigid company structure? Upload an example document and input a follow-up query.
MCP gives AI the keys to your data
That example illustrates the crux of the issue: If generative AI acts as a metaphorical brain, how do we give it access to the data we need to process? How do we connect generative AI to a body of tools to accomplish downstream tasks? The solution is Model Context Protocol (MCP). MCP is an open-protocol model built by Anthropic that provides a system designed to link advanced LLMs to data across an organization.5 This allows advanced AI models to communicate with and leverage company data in a standardized manner, eliminating the need for developers to teach the models how to communicate with different sets of data. In simple terms? It helps everyone speak the same language.
This creates a “plug-and-play” scenario for AI and an organization's data, helping to dissolve data silos and unleash a host of tools for AI to leverage in an increasingly autonomous role. If these “agents” were colleagues, they would be harnessing their inner Daft Punk to work “harder, better, faster, stronger” to drive streamlined organizational success. However, too many options can become problematic. In the words of the late Steve Jobs: “Innovation is not about saying yes to everything. It’s about saying no to all but the most crucial features.” This means the adoption of AI must solve a specific problem and/or an unmet organizational need. To this end, MCP is unlocking a new level of modularity that form-fits AI to a company’s specific goals.
Rethinking AI’s cognitive trade-offs
Despite the fervent enthusiasm surrounding AI, one must address the woolly mammoth in the room. Recent research has shown an inverse correlation between self-reported confidence in generative AI and critical thinking when engaging in AI-assisted tasks.6 As user confidence in GenAI increases, critical thinking during the utilization of AI decreases. However, users with high task-specific self-confidence bucked that trend by applying more critical thinking during their utilization of AI.6 A larger and more diverse study showed a negative correlation between frequency of AI usage and critical thinking skills.7 This correlation was stronger in younger people, while higher educational level was correlated with greater critical thinking, regardless of AI usage. These studies raise the issue of a hypothetical feedback loop: reliance on AI decreases critical thinking skills, which increases reliance on AI, which decreases critical thinking skills, and so on and so forth—in perpetuity.6,7
For broader perspective on the potential impact of newer technologies on humans’ cognitive ability, there are numerous studies associating increased utilization of GPS with decreased spatial awareness and cognitive mapping.8,9 On the contrary, older neuroimaging studies (circa 2000–2011) on taxi drivers manually learning the roads of London demonstrated physiological changes in a region of the brain associated with spatial memory.10-12 Together these studies support the “use it or lose it” principle of cognitive function and underscore the dual benefit-drawback nature of modern technology.
The coexistence of AI and critical thinking
The AI studies demonstrate correlation, not causation. An alternative interpretation of the Lee study6 is that workers who use GenAI but have less self-reported confidence in it have a better understanding of the limitations of AI. Therefore, they take more time to put in the effort to think critically about its output. This explanation also aligns with the positive correlation between self-confidence in a task and critical thinking while using AI, because these workers likely have manually completed the task frequently in the past, thus giving them a better understanding of what a quality output should look like.
AI and critical thinking are not at odds; rather, they hold the potential to synergize and amplify human potential.
However, this leaves us with an important question: As AI continues to advance in capabilities, can this increased cognitive offloading be outweighed by the benefits of freeing up the brain to focus on more creative and complex tasks? After all, Steve Jobs wore the same outfit—black turtleneck, jeans, and New Balance sneakers—every day to reduce decision fatigue. And it’s safe to say that we all know how that turned out.
Human ingenuity is at the heart of AI
Streamlining corporate operations requires more than strategic implementation of AI. Instead, it calls for a parallel investment of resources that train staff on the very processes that AI will augment. By seamlessly combining this human skill and creativity with a focused, purpose-driven application of AI, organizations have the potential to redefine success within their respective industries. AI—albeit flashy—is simply another tool for humans to shape the world around us. After all, in another 300,000 years, humans undoubtedly will look back and chuckle, marveling at how we were still only using “stone tools.”
Ready to stay at the forefront of pharmaceutical marketing? Discover what we can do for you at DeerfieldGroup.com/contact-us.
References: 1. Handwerk B. An evolutionary timeline of Homo sapiens. Smithsonian Magazine. https://www.smithsonianmag.com/science-nature/essential-timeline-understanding-evolution-homo-sapiens-180976807/. Published February 2, 2021. Accessed May 21, 2025. 2. Wood B. Colloquium paper: reconstructing human evolution: achievements, challenges, and opportunities. Proc Natl Acad Sci USA. 2010;107 Suppl 2(Suppl 2):8902-8909. doi:10.1073/pnas.1001649107. 3. Wong K. How scientists discovered the staggering complexity of human Evolution. Scientific American. Published online September 2020. doi:10.1038/scientificamerican0920-66. 4. Neser L. Introduction to Earth science, second edition. Published online January 30, 2025. doi:https://doi.org/10.21061/introearthscience2e. 5. Anthropic. Introducing the Model Context Protocol. Anthropic.com. https://www.anthropic.com/news/model-context-protocol. Published 2024. Accessed May 21, 2025. 6. Lee H, Sarkar A, Tankelevitch L, et al. The impact of generative AI on critical thinking: self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. Published online 2025. doi:10.1145/3706598.3713778. 7. Gerlich M. AI Tools in society: impacts on cognitive offloading and the future of critical thinking. Societies. 2025;15(1):6. doi:10.3390/soc15010006. 8. Clemenson GD, Maselli A, Fiannaca AJ, et al. Rethinking GPS navigation: creating cognitive maps through auditory clues. Scientific Reports. 2021;11(1). doi:10.1038/s41598-021-87148-4. 9. Dahmani L, Bohbot VD. Habitual use of GPS negatively impacts spatial memory during self-guided navigation. Scientific Reports. 2020;10(1). doi:10.1038/s41598-020-62877-0. 10. Maguire EA, Gadian DG, Johnsrude IS, et al. Navigation-related structural change in the hippocampi of taxi drivers. Proceedings of the National Academy of Sciences. 2000;97(8):4398-4403. doi:10.1073/pnas.070039597. 11. Maguire EA, Woollett K, Spiers HJ. London taxi drivers and bus drivers: A structural MRI and neuropsychological analysis. Hippocampus. 2006;16(12):1091-1101. doi:10.1002/hipo.20233. 12. Woollett K, Maguire Eleanor A. Acquiring “the knowledge” of London’s layout drives structural brain changes. Current Biology. 2011;21(24):2109-2114. doi:10.1016/j.cub.2011.11.018.
Acknowledgements:
Trevor Fusaro, Juan Vasquez, and Aiden Muneath for thoughtful discussions and suggestions.
Samantha Pollock for copy support.
