In a recent edition of Eye on AI, various developments in the field were highlighted, including a study indicating that artificial intelligence (AI) can effectively function as a team player. OpenAI has promoted its Chief Operating Officer, while CEO Sam Altman is redirecting his focus. Meanwhile, Apple has made changes to its AI team due to delays in Apple Intelligence features. Additionally, a groundbreaking AI-based weather forecasting method has been introduced and AI is also transforming the field of architecture.
Research continues to underscore the positive effects of AI on productivity. While many executives initially view AI as a possible substitute for human labor, seeing AI as a tool to fully automate tasks and reduce headcount, data suggests this may not be the most effective way to perceive the technology. There are instances where AI can completely automate tasks; however, most current AI systems, including the AI agents from Salesforce, ServiceNow, Microsoft, and Google, have yet to reach a level of capability and reliability for full task automation. Instead, these AI systems are better viewed as complementary to human labor, enhancing human performance rather than replacing it.
This perspective is supported by a recent study conducted by researchers from prestigious institutions including Harvard, the Wharton School of Business at the University of Pennsylvania, ESSEC Business School in France, and consumer products giant Procter & Gamble. This research was published as a working paper on the SSRN research repository. One of the authors, Ethan Mollick from Wharton, is notable for his insights on using AI effectively in business.
In 2024, a one-day virtual product development workshop at Procter & Gamble resembled the company’s traditional process but incorporated an AI element. The focus was on the “seed” stage of product development, which entails brainstorming potential new product ideas and developing them to decide on further testing. Normally, Procter & Gamble uses two-person teams, combining Commercial operations personnel with R&D experts for brainstorming. The researchers in this study involved 776 P&G employees, who were randomly assigned to work alone, work alone with a generative AI assistant based on OpenAI’s GPT-4 model, work in two-person teams, or work in two-person teams with AI assistance.
The groups were tasked with generating new consumer product ideas for P&G’s divisions, including baby care, feminine care, grooming, and oral care. These ideas were evaluated by human judges with business and technology expertise.
The findings revealed that AI significantly boosts individual performance. Individuals working alone without AI assistance performed the worst, while those using AI outperformed two-person teams without AI. The performance of AI-assisted individuals was not statistically different from two-person teams with AI, suggesting that AI can serve as an effective substitute, enabling a reduction in team sizes to single individuals working with AI.
Other advantages for individuals using AI included faster work, with over 16% less time needed to develop an idea compared to those without AI, and teams using AI worked about 12% faster. AI assistance also improved emotions during the brainstorming process and led to a more integrated approach in idea generation, blending technical and commercial innovations.
However, the most intriguing finding was that two-person teams with AI produced a higher number of ideas deemed “exceptional” by human judges—ideas considered most likely to result in groundbreaking products. Teams assisted by AI also enjoyed the process more compared to other groups.
Ethan Mollick commented that organizations often see AI as just another productivity tool but emphasize that it offers potential for critical thinking and complex problem solving. He highlighted that AI could serve as a collaborator, not merely a tool, encouraging organizations to reimagine work structures rather than focusing solely on automation. Mollick noted that while the cost-saving benefits of AI are tempting, using AI to empower people can lead to extraordinary outcomes.
The article concludes with an update on other AI-related news.
