COBOL migration where a difference of even 1 byte results in failure: what is the accuracy rate of OpenAI's new model?
Key point
To what extent can AI handle COBOL and Fortran, which have few learning resources? OpenAI announced that in a test constructing database engines in 20 languages, the average pass rate of the new model increased from the conventional model's 26.8% to 87.4%.
Attention is being focused on how far AI can handle COBOL and Fortran, which have few learning resources. OpenAI announced that in a test constructing database engines in 20 languages, the average pass rate of the new model increased from the conventional model's 26.8% to 87.4%.
Results are also being sought regarding the evaluation of rewriting existing COBOL programs into other languages.
This item is a summary. The full, checkable version is at the original source:
Read the original at ITmedia AI+ ↗