This blog post examines the unique development patterns of computer science by comparing them to Kuhn and Popper’s theories of scientific revolution.

Introduction

Computer science is a discipline with a short history, yet it has developed through experiencing extremely rapid change. Computer science has achieved rapid development across various fields within a short timeframe, and its pace of development continues today. For example, the 2005 Intel Xeon 3.6GHz boasts 6,505 times the performance of the 1978 VAX-11/780. Recently, as CPU clock speeds reached their limits, the paradigm shifted to multicore processors. Beyond this, the integration of diverse technologies keeps computer science in a state of constant change, and the impact of these changes on modern life is profound.
As recently as the 1940s, computers were enormous machines filling entire rooms. Today, however, the proliferation of smartphones has ushered in an era of “one computer per person,” enabling us to accomplish numerous tasks. Unlike other scientific disciplines, computer science has advanced through performance improvements and paradigm shifts, a phenomenon noteworthy from the philosophical perspective of scientific revolution. We will examine the principles and direction of computer science’s development, focusing on the theory of scientific revolution.

Thomas Kuhn and Karl Popper’s Arguments on Scientific Revolution

Scientific revolutions profoundly impact not only academia but society as a whole. Thomas Kuhn described scientific revolution as a ‘paradigm shift,’ defining a paradigm as a scientific achievement capable of posing significant problems for researchers. According to Kuhn, while a paradigm is effectively functioning, existing theories are extended and clarified, enabling scientists to engage in normal science. However, when the existing paradigm fails to adequately explain natural phenomena, a new paradigm emerges, leading to a scientific revolution.
Karl Popper argued that science progresses not through the verification of hypotheses, but through falsification. He viewed scientific discoveries as non-definitive, believing no hypothesis could be 100% certain. Popper argued that scientific truth must be pursued through constant doubt and criticism, asserting this process is the essence of scientific progress. This philosophy of Popper’s is known as falsificationism.

The Development Principle of Computer Science from the Perspective of the Scientific Revolution

Computer science is a discipline that primarily develops with the goal of improving performance. New methodologies to maximize performance are continually proposed across various layers, including hardware devices, operating systems, and application software. In this process, correctness is as important as performance. In computer science, correctness is a crucial criterion for determining whether a system operates accurately, and there exists a philosophy that performance, no matter how high, must not deviate from correctness.
Unlike other sciences, computer science sees new paradigms emerge very rapidly, and their validation also occurs quickly. For example, while sequential execution of CPU instructions guaranteed correctness, it was inefficient. This led to the introduction of new methodologies like pipelined execution and out-of-order execution. While these methods aim for performance gains, solutions to maintain correctness throughout the process must also be provided.

Modern Developments in Computer Science: The Rise of Artificial Intelligence and Quantum Computing

Another crucial advancement in modern computer science is the rise of Artificial Intelligence (AI) and Machine Learning (ML). In recent years, AI has evolved from simple data processing into a technology enabling complex tasks like autonomous driving, natural language processing, and image recognition. This is not only enhancing computer science performance but fundamentally transforming human-machine interaction. In particular, the advancement of deep learning technology provides the ability to rapidly process large volumes of data and identify complex patterns within it, thereby solving problems that are difficult for humans to resolve.
Furthermore, quantum computing presents a new paradigm that transcends traditional binary logic-based computing methods.
Quantum computers possess the potential to rapidly process complex problems unsolvable by conventional computers, and this is expected to drive innovation across diverse fields such as cryptography, chemical simulations, and optimization problems. These technologies are currently leading the advancement of computer science and are being recognized as crucial elements that will drive future paradigm shifts.

Differences Between Scientific Revolutions and Computer Science

While the development trajectory of computer science shares similarities with traditional scientific revolution theories, distinct differences exist. In Kuhn’s scientific revolutions, paradigm shifts primarily occur when existing theories reveal limitations and are replaced by new ones. In contrast, computer science often improves performance or proposes new directions while maintaining the existing paradigm. Computer science is a discipline where new methodologies emerge rapidly based on necessity rather than building solely on past achievements. Consequently, the traditional framework of scientific revolution struggles to fully explain its developmental process.
Computer science evolves by constantly balancing performance enhancement and correctness. Furthermore, the advent of mobile environments and the rise of new paradigms like artificial intelligence demonstrate that computer science is a discipline developing simultaneously in multiple directions. In this regard, computer science requires a more flexible theory of scientific revolution.

Conclusion

Computer science is a discipline that has developed through rapid change within a very short timeframe. Its evolution cannot be fully explained by the existing framework of scientific revolutions alone. Computer science, where the balance between performance enhancement and correctness, along with the introduction of new paradigms, is continuously being pursued, can be seen as one of the key cases for the theory of scientific revolutions. Its potential for future development remains boundless.