Meta has acknowledged that its first attempt at creating a 'superintelligence' failed to meet expectations, lasting only three days before it was deemed ineffective. The company has been exploring advanced artificial intelligence systems, aiming to push the boundaries of machine learning and cognitive capabilities.

This early version of superintelligence reportedly struggled with basic tasks and did not demonstrate the advanced reasoning abilities that were anticipated. Meta's team is now reassessing their approach to developing AI technologies, focusing on improving the foundational elements necessary for more robust and reliable systems in the future.