What: The BCG reviews how AI is used in the corporate environment.

Why it is important: AI is a tool, not an endpoint, and might create a loss of creativity within organizations.

In an experiment with over 750 BCG consultants, people used generative AI (GPT-4) very effectively for creative ideation, boosting performance by 40%. But for complex business problem solving, a task outside GPT-4's competence, relying on its output reduced performance by 23%.

People seem to mistrust AI where it can add value, but trust it too much where it lacks competence. This held true even when participants were warned of GPT-4's limitations.

Using GPT-4 led to a 41% drop in diversity of ideas, as people converged on the AI's repetitive outputs. 70% of participants worried relying on AI may reduce their own creativity over time.

To benefit from generative AI, companies need an advanced data strategy, new roles/workflows that focus human effort on the remaining uniquely human tasks, strategic workforce planning to cultivate diversity of thought, and rigorous testing and experimentation as AI capabilities rapidly advance.

Generative AI adoption poses a double-edged sword - tremendous gains in efficiency but also risks around overreliance. Leaders must approach it as a massive change management challenge.

How people create (and destroy) value thanks to AI