Having a human in the loop is critical to mitigating the risks of generative AI errors and biases. But humans are also vulnerable to errors and biases and may trust artificial intelligence either too much or not enough. Findings from a field experiment by MIT and Accenture suggest that targeted friction in the form of labels that flag potential errors and omissions can direct users' attention to content that should be given closer inspection without sacrificing efficiency.

Artificial intelligence is transforming all sectors of the economy, but there's no reason to fear that robots will replace all human employees. In fact, companies that automate their operations mainly to cut their workforces will see only short-term productivity gains, say the authors. Their research, involving 1,500 firms in a range of industries, shows that the biggest performance improvements come when humans and smart machines work together, enhancing each other's strengths. People need to train AI agents, explain their outputs, and make sure they are used responsibly. AI agents, in turn, can assist people with information gathering, data crunching, routine customer service, and physical labor, thereby freeing them for higher-level tasks that require leadership, creative thinking, judgment, and other human skills. To get the most out of AI, companies need to redesign their business processes. After deciding what needs improvement--their operational flexibility, speed, or scalability; their decision making; or their ability to personalize products and services--they can devise appropriate solutions. That will mean not only implementing AI technology but also developing employees who can work effectively at the human-machine interface. The authors describe how a number of firms are already taking these steps and optimizing collaborative intelligence. But many more should follow their example.

This is an MIT Sloan Management Review article. A new global study finds several new categories of human jobs emerging. These roles are not replacing old ones. They are brand-new positions that complement the tasks performed by AI machines and will require skills and training that have never before been needed.