Computer simulation (CS) and thought experiments (TE) seem to produce knowledge about the world without intervening in the world. This has called for a comparison between the two methods. However, Chandrasekharan et al. (2013) argue that the nature of contemporary science is too complex for using TEs. They suggest CS as the tool for contemporary sciences and conclude that it will replace TEs. In this paper, by discussing a few TEs from the history of science, I show that the replacement thesis about TE is a failure. The paper is divided into three sections. The first section discusses the arguments of Chandrasekharan et al. (2013) and demonstrates the three distinct aspects of the replacement thesis. The second section examines the argument against TE and shows that they are inadequate to prove the withering of TE from science. The third section discusses Albert Einstein’s Magnet and Conductor TE and demonstrates that replacing such TE with CS yield no advantage.

计算机模拟（CS）和思想实验（TE）似乎可以在不干预世界的情况下产生有关世界的知识。这要求在两种方法之间进行比较。但是，Chandrasekharan等。（2013年）认为，当代科学的本质对于使用TE来说太复杂了。他们建议使用CS作为当代科学的工具，并得出结论，它将取代TE。在本文中，通过讨论一些科学史上的TE，我证明了关于TE的替代论文是失败的。本文分为三个部分。第一部分讨论Chandrasekharan等人的论点。（2013年），并论证了替代论文的三个不同方面。第二部分研究了反对TE的论点，并表明它们不足以证明TE的科学衰落。