Back to today’s briefing

Story perspectives

AI Falls Short in Simulated Workplace Performance

4/27/2025

25 6

Summary

Story summary

In a revealing experiment at Carnegie Mellon University, AI agents struggled in a simulated software company, with the top performer managing just 24% of tasks. This stark performance highlighted AI's shortcomings in common sense and social skills, underscoring that these technologies are far from ready to take on intricate human roles.

1 of 1

Story summary

In a revealing experiment at Carnegie Mellon University, AI agents struggled in a simulated software company, with the top performer managing just 24% of tasks. This stark performance highlighted AI's shortcomings in common sense and social skills, underscoring that these technologies are far from ready to take on intricate human roles.