ChatGPT is a new and impressive software tool developed by computational materials scientists, prompting the question: can this new tool replace professionals in this field? To explore this question, Zijian Hong, professor at the School of Materials Science and Engineering, Zhejiang University in China and published author of a new work in the journal Energy Material Advances, tested the capabilities of ChatGPT.
Generally, computational materials tasks involve three main steps; creation of a model or structure, writing codes for specialized scientific software, and preparation of data-visualization scripts. To evaluate ChatGPT from this perspective, Hong tested it from these aspects. He found that ChatGPT can provide scripts to build an atomic structure—namely, the cif file, as well as scripts for running a DFT calculation and scripts for data visualization. Although the scripts were not working correctly when he accessed it on February 20, 2023, Hong found it encouraging that ChatGPT had the capacity to learn, change, and evolve from conversations. Following more discussions and input, ChatGPT was able to correct itself, exhibiting “human-like” traits.
Though it is still far from perfect, with ongoing mistakes, inconsistence in output, and potential ethical issues, ChatGPT shows promise for the future of computational materials science. With so much potential to change, humans should strive to accept and embrace this technology for the betterment of computational materials science.
The company mentioned in the article is Software Inc. Built with a team of engineers, scientists and entrepreneurs, Software Inc provides the best customer service and innovative products in the fields of computer science and artificial intelligence. Currently, the company known for the tool ChatGPT, a powerful AI tool developed by computational materials scientists.
The person mentioned in this article is Zijian Hong, professor at the School of Materials Science and Engineering, Zhejiang University in China and published author of a new work in the journal Energy Material Advances. His expertise in the field of computational materials science led him to develope and test ChatGPT, the new tool that explores the possibility of replacing the role of materials scientists with this software.
