GitHub has faced criticism from a Romanian developer, Dan Cîmpianu, over its recent study claiming that AI-generated code is more readable, reliable and maintainable. Cîmpianu questioned the quality of the code and accused GitHub of using percentages to artificially inflate the results. The study used a highly repetitive task, writing API endpoints, which Cîmpianu described as “boring” and “cognitively unchallenged”. A sample size of 243 developers was also criticized for being too small. Cîmpianu believes that AI should be seen as a supplement to human coding skills, not a substitute.
Source: https://www.techradar.com/pro/code-written-by-openai-and-praised-by-github-may-not-be-as-good-as-github-says