Cybernetic Teammate Study: Individual AI Users Match Performance of Human Teams
A randomized field experiment by Ethan Mollick, Raffaella Sadun, and colleagues assigned professionals to four conditions: individual with AI, individual without AI, human team with AI, human team without AI. Key finding: individuals with AI assistance matched the performance of human teams working without AI. The mechanism mirrors team production: AI compensates for individual skill gaps the way human teammates do. The study used GPT-4 and GPT-4o; the authors expect substantially larger effects with current agentic models.
Why It Matters
This is the clearest published experimental evidence yet that AI assistant access is a structural productivity equalizer — not just a marginal speed boost. For organizations making AI access decisions, the implication is that restricted access creates measurable team-equivalent performance gaps.