Get the latest tech news
AI models still struggle to debug software, Microsoft study shows
Even some of the best AI models today still struggle to debug software, a Microsoft study shows.
The study’s co-authors tested nine different models as the backbone for a “single prompt-based agent” that had access to a number of debugging tools, including a Python debugger. According to the co-authors, even when equipped with stronger and more recent models, their agent rarely completed more than half of the debugging tasks successfully. Many studies have shown that code-generating AI tends to introduce security vulnerabilities and errors, owing to weaknesses in areas like the ability to understand programming logic.
Or read this on TechCrunch
