The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and Expertise
Study done by P&G testing randomized AI usage with professionals
Hello Curious Minds Crew,
As I told my son recently regarding learning and using AI tools, “Fools don’t use available tools. We have new and emerging tools that prehistoric cavemen didn’t have, don’t be a caveman.” LOL no offense to cavemen.
Proctor & Gamble (P&G) did a randomized study with a set of professionals where some used AI and some didn’t, and some worked individually and others worked in teams. Here is the short summary overview:
Summary of Key Points: The report details a field experiment conducted at Procter & Gamble (P&G) to study the impact of Generative AI (GenAI) on teamwork and expertise. The study involved 776 professionals working on real product innovation challenges, randomly assigned to work individually or in teams, with or without AI assistance (GPT-4). The research aimed to understand how AI affects performance, expertise sharing, and social engagement in collaborative work. Key findings include:
Performance Enhancement: Individuals using AI achieved performance levels comparable to teams without AI, suggesting AI can effectively substitute for certain benefits of human collaboration. AI also significantly reduced the time spent on tasks while increasing the length and detail of solutions.
Expertise Broadening: AI helped break down functional silos. Professionals with AI produced balanced solutions regardless of their background (R&D or Commercial), whereas those without AI tended to produce solutions aligned with their expertise. AI allowed less experienced employees to achieve performance levels comparable to teams with experienced members.
Social Engagement: Participants using AI reported more positive emotional responses (enthusiasm, energy) and fewer negative emotions (anxiety, frustration) compared to those working alone without AI. This suggests AI can fulfill some of the social and motivational roles traditionally offered by human teammates.
Exceptional Performance: AI-augmented teams were more likely to generate solutions ranked in the top 10%, indicating AI's potential to drive breakthrough innovations.
AI Usage Patterns: Participants showed high retention rates of AI-generated content, but the semantic fingerprint of solutions remained closer to human-only solutions than pure AI outputs, indicating meaningful human shaping and contextualizing of AI suggestions.
Full report: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5188231
There is a saying floating around that goes something like this, “AI won’t replace insurance agents, but agents who don’t use AI will get replaced by agents who do!” I think this mantra has merit. Live a life of endless learning, always leaning into curiosity, and implementing/executing on what you learn.
Cheers to the Curious Minds Crew! Until next time, stay curious.
TV