In this study conducted by Boston Consulting Group and the researchers, the performance implications of Artificial Intelligence (AI) on complex and knowledge-intensive tasks are examined. The experiment involved 758 consultants, and they were randomly assigned to one of three conditions: no AI access, GPT-4 AI access, or GPT-4 AI access with a prompt engineering overview. It is observed that AI can significantly improve productivity and quality in tasks within its capabilities, with consultants completing more tasks quickly and producing higher-quality results. However, for tasks outside the capabilities of AI, consultants using AI were less likely to produce correct solutions. The study also identifies two patterns of human-AI integration – “Centaurs” and “Cyborgs.”
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4573321