Paper Writing Skill Test of ChatGPT for Causal Inference of Physical AI with LMM

Abstract: This paper deals with application ChatGPT for writing engineering paper. Basically, we have to write paper from bottom to final paper such as reference, summarizing, a title selection, and writing work including English skill reviewing. However, the many apples like ChatGPT provide many functions as well as information for producing paper, report, research plan, and etc. This paper focuses physical AI including causal inference and tries its paper writing skill for user for physical causal inference. Causal inference is a big issues and problem in physical AI and LLM (Large Language Model), and LMM (Language Multi-Model) to operate independently in Internet in industrial or small device or robot. The Physical AI have to use LMM (Language Multi-Model) with language sensor, several image sensors, temperature sensor, and others for operation through the communication with user. LMM (Language Multi-Model) such as ChatGPT or AI models has the significant problems in application because of the deficit causal inference of function language recognition, image, audio model, and etc.

That is, the causal inference problems of the large Language Multi-Model (LMM) remain issue for application of physical AI. Because the causal problems are difficult to describe reason in natural language model, this function restricts the application of physical AI in industrial fields. This paper selects this title and to see how they (ChatGPT) can write causal inference issues of this physical AI to see improving skill of ChatGPT for the causal ability of LMM. Of course, there are many limitations in writing paper by ChatGPT and we cannot believe fully the results of ChatGPT. However, this paper tries how to use and how much it is true in the results. In the future, we absolutely need experiment and compare.

Keywords: LMM, ChatGPT, Causal Inference, Physical AI, Robot.