Sandeep Ozarde
Young children outperform large language models in problem-solving tasks that require innovation. Children ages three to seven displayed an 85% success rate in choosing the correct object to complete a task without the proper tool, while large language models achieved only 76% accuracy. The findings highlight the lack of ability of large language models to introduce novel ideas and emphasize the importance of embodying active, curious intervention in AI systems to foster innovation.
0 Comments 0 Likes
App Store
Download Artifact to read and react to more links
App Store Play Store