Abstract
In this study, training and prediction performance on turning data were investigated with ChatGPT, which is a popular AI platform nowadays. In this context, the resultant cutting forces obtained as a result of different turning simulations with FEM, training and estimation were made using regression, ANN, ANFIS methods. Using the same data, training and predictions were made with ChatGPT-3 with different prediction algorithms. As a result, the lowest average error rates in the predictions made with the training data; 2E-6% for ANFIS in prediction methods and 0.19% for ANN1 conversation in GPT-3 were obtained. The lowest average error rates in the predictions made with the test data; 5.41% for regression using logarithmic Box–Cox transformation in prediction methods, and 22.66% for ANN1 conversation in GPT-3 were achieved. The highest prediction performance in GPT-3 conversations was observed when GPT-3 was asked to make predictions with ANN algorithm on both training and test data. As a result, GPT-3 has not yet generated acceptable solutions for machining problems due to its low performance in predicting test data. However, due to the fast advancement of artificial intelligence technologies, it is obvious that solutions to this and more engineering problems will be generated in near future.
Highlights
Prediction with ChatGPT-3
Prediction performance of artificial intelligence
Cutting force prediction of turning operation
Acknowledgements
The author acknowledges that the training and prediction of the datasets in this article were generated in part by ChatGPT (powered by OpenAI's language model GPT-3; http://openai.com). Editing was done by the author.
Author contribution
The author confirms sole responsibility for the following: study conception and design, data collection, analysis and interpretation of results, and manuscript preparation.
Data availability statement
The data generated and/or analyzed during the current study are not publicly available, but data may be provided by the corresponding author upon reasonable request.
Disclosure statement
No potential conflict of interest was reported by the author(s).