In a recent edition of “Eye on AI,” multiple developments in the AI sector were discussed, including a new study that highlights AI’s potential as a collaborative tool rather than just an automating force. OpenAI is set to promote its COO, while CEO Sam Altman adjusts his focus. Meanwhile, Apple is reorganizing its AI team amidst delays in their Apple Intelligence features. A groundbreaking AI-based weather forecasting method has emerged, and AI is also making waves in the field of architecture.
As evidence accumulates regarding AI’s positive effect on productivity, a distinction is emerging in how AI should be understood by businesses. Although some executives consider AI to be a full substitute for human labor with the aim of automating tasks completely, recent data contradicts this assumption. Current AI systems, such as those from companies like Salesforce, ServiceNow, Microsoft, and Google, are not yet consistently capable of performing tasks autonomously. Instead, AI is more effectively viewed as enhancing human labor, improving overall workforce performance rather than replacing it.
This perspective is reinforced by a new study from researchers associated with Harvard, the Wharton School of Business at the University of Pennsylvania, ESSEC Business School in France, and Procter & Gamble. Published as a working paper on SSRN, the research involved a virtual product development workshop at P&G in 2024. The workshop incorporated an AI-based approach to the initial “seed” stage of product development, requiring P&G employees to brainstorm new product ideas. Participants included 776 employees from Commercial and R&D sectors, randomly assigned to work alone, with access to a generative AI assistant using OpenAI’s GPT-4 model, or in traditional two-person teams with or without AI assistance.
The study found that individuals working independently but with AI assistance outperformed those in two-person teams without AI. These AI-assisted individuals were also not statistically outperformed by the two-person teams who used AI. A notable advantage of AI was its ability to help individuals work faster, reducing time by over 16% compared to those working without AI, although two-person teams with AI were 12% faster as well.
AI demonstrated additional benefits, as individuals reported more favorable emotions during the process than those working alone without AI. Alone, participants tended to generate ideas aligned with their professional backgrounds, but with AI, they achieved more comprehensive approaches, blending commercial and technical innovations.
However, despite the findings, the study highlights the substantial benefits of maintaining team sizes. Two-person teams using AI produced more standout ideas, deemed by human experts as “exceptional” and likely to lead to breakthrough products. Additionally, these teams reported the highest levels of job satisfaction.
Ethan Mollick, one of the researchers, blogged that AI should be considered as a team collaborator rather than merely a productivity tool. He suggested that organizations should reimagine work structures, focusing beyond mere efficiency gains from AI adoption, and instead fostering workplaces where AI is used to enhance human capabilities.
Jeremy Kahn of Fortune further elaborated on these developments, emphasizing the tendency of some managers to prioritize immediate labor savings over long-term strategic benefits associated with AI-enhanced collaboration.
In related news, OpenAI’s organizational restructure and Apple’s challenges with its AI features illustrate the dynamic and evolving landscape of AI integration across industries.