Arthur Besse to [email protected]English • 2 months ago"participants who had access to an AI assistant wrote significantly less secure code" and "were also more likely to believe they wrote secure code" - 2023 Stanford University study published at CCS23arxiv.orgmessage-square38fedilinkarrow-up1370arrow-down110
arrow-up1360arrow-down1external-link"participants who had access to an AI assistant wrote significantly less secure code" and "were also more likely to believe they wrote secure code" - 2023 Stanford University study published at CCS23arxiv.orgArthur Besse to [email protected]English • 2 months agomessage-square38fedilink
minus-square@Hasherm0nlink7•2 months agoUntil you find out those were also built by a junior using an llm to help 🙃
Until you find out those were also built by a junior using an llm to help 🙃