Date context: institutional-research-summary
HBS describes an experiment with product-development tasks at Procter & Gamble. Individuals using AI reached the study's quality level of two-person teams without AI. Solutions were less divided by participants' commercial or technical specialisms. The research raises questions about sharing knowledge across functions.
Dell'Acqua et al.; Harvard Business School AI Institute / NBER · 2025-04-02
HBS institutional research summary checked; full experiment and methodology not independently reviewed.
What the source shows
HBS describes an experiment with product-development tasks at Procter & Gamble. Individuals using AI reached the study's quality level of two-person teams without AI. Solutions were less divided by participants' commercial or technical specialisms. The research raises questions about sharing knowledge across functions. Source: Dell'Acqua et al.; Harvard Business School AI Institute / NBER, 2025-04-02
What the material does not establish
Results concern specific innovation tasks at one company. They do not show that AI replaces teams, trust or human relationships. The institutional summary was read; linked full versions could not be fetched, so the complete experimental method was not reviewed here. Source: Dell'Acqua et al.; Harvard Business School AI Institute / NBER, 2025-04-02
For HR, AI + JJ
AI + JJ, application: Combine manager, employee and service perspectives on a fictional onboarding problem. Does AI help colleagues understand each other's contributions, and improve the solution under joint review? Source: Dell'Acqua et al.; Harvard Business School AI Institute / NBER, 2025-04-02