A ZDNET writer tested Perplexity’s agentic AI system by assigning it five complex tasks to execute on a Mac computer. The testing resulted in a positive assessment, with the writer indicating sufficient confidence in the system’s performance to use it again for similar assignments.