Turing test for code-generation tools
A human judge examines the code generated by one programmer and one code-generation tool for the same formal specification. If the judge cannot reliably tell the tool from the human, the tool is said to have passed the test
Recent Comments