Understanding AI’s Strength in Normal Science
Artificial intelligence capabilities have made significant strides in recent years, enabling machines to translate languages, analyze massive datasets, and even surpass human performance on select scientific tasks. However, according to Keith McCormick, the 2025–2026 Bodily Bicentennial Professor of Analytics at the University of Virginia Darden School of Business, there are clear boundaries to what AI can accomplish. In a recent presentation, McCormick shared his insights on where artificial intelligence truly shines—and where it still falls short.
The Distinction Between Normal Science and Breakthroughs
McCormick draws on the work of renowned science philosopher Thomas Kuhn, who introduced the concept of “normal science.” Normal science refers to the day-to-day scientific activities that operate within established frameworks, involving incremental improvements and problem solving. According to McCormick, artificial intelligence capabilities are perfectly suited to these tasks. AI can process vast amounts of data, identify patterns, and optimize solutions within known parameters. These strengths have led to remarkable progress in fields such as language translation, data analytics, and even certain areas of scientific research.
But McCormick warns that, despite the impressive advancements in AI, the technology is unlikely to deliver the major paradigm-shifting breakthroughs that define scientific revolutions. Such breakthroughs, he argues, still rely heavily on human creativity, intuition, and the willingness to challenge existing assumptions.
Current Boundaries of Artificial Intelligence
AI’s rapid development has inspired speculation about the arrival of artificial general intelligence (AGI)—machines capable of understanding and performing any intellectual task that a human can. Yet, McCormick contends that while the artificial intelligence capabilities we see today are powerful, they are fundamentally constrained by their reliance on existing data and predefined rules. AI systems are excellent at optimizing known processes but lack the ability to conceive truly original ideas or theories.
This limitation becomes particularly evident when considering the nature of scientific breakthroughs. Historically, revolutionary advancements—such as the discovery of relativity or quantum mechanics—have emerged from the ability to see beyond current paradigms and embrace novel perspectives. McCormick suggests that, until AI can transcend its current framework, these transformative leaps will remain the domain of human thinkers.
The Role of Low-Code Tools in Expanding AI Access
One area where AI is making a substantial impact is in the democratization of data science through low-code and no-code platforms. McCormick highlights the hidden value of these tools, which allow users with limited programming experience to build predictive analytics models and harness artificial intelligence capabilities for business and research purposes. By lowering the technical barriers, these platforms are enabling a broader range of professionals to participate in data-driven innovation.
Nevertheless, McCormick cautions that while low-code tools expand access, they do not eliminate the need for critical thinking and domain expertise. Understanding the context and limitations of AI-driven insights remains essential for avoiding common pitfalls, such as misinterpreting data or over-relying on automated solutions.
AI’s Impact on Scientific and Business Practice
The practical applications of artificial intelligence capabilities are already reshaping numerous industries. From healthcare diagnostics to financial forecasting and supply chain optimization, AI is delivering efficiency gains and uncovering new opportunities. McCormick, who has spent over thirty years building predictive analytics models across various sectors and has authored numerous books on the subject, emphasizes the importance of using AI responsibly and understanding both its strengths and its boundaries.
As organizations integrate AI into their operations, McCormick encourages leaders to focus on blending machine efficiency with human creativity. By leveraging AI for routine analytical tasks while leaving room for human-driven innovation, businesses can maximize their potential for both incremental improvements and transformative breakthroughs.
The Future: AI and Human Collaboration
Looking ahead, McCormick believes that the most promising path forward lies in fostering collaboration between humans and machines. While AI will continue to excel at normal science and data-intensive problem solving, humans will remain essential for pioneering new ideas and challenging established paradigms. This symbiotic relationship, he argues, will drive progress in both science and business.
In conclusion, the current state of artificial intelligence capabilities is remarkable, but the next wave of scientific revolutions will likely depend on human ingenuity. As AI tools become more accessible and powerful, the challenge will be to harness their strengths while preserving the uniquely creative spark that leads to true breakthroughs.
This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.