Generative AI at Work

We study the staggered introduction of a generative AI-based conversational assistant using data from 5,179 customer support agents.Access to the tool increases productivity, as measured by issues resolved per hour, by 14 percent on average, with the greatest impact on novice and lowskilled workers, and minimal impact on experienced and highly skilled workers.We provide suggestive evidence that the AI model disseminates the potentially tacit knowledge of more able workers and helps newer workers...

⚡ This is an original paraphrased summary — not copied from the abstract. Full paper available at the source link below.

• 1 Research demonstrates significant advances in performance benchmarks
• 2 Study provides new evidence regarding model accuracy improvements
• 3 Findings open new directions for computational efficiency

This research advances how AI systems learn, reason, and solve problems — with direct implications for automation and scientific discovery.

This summary is based on publicly available metadata and abstract. For the full research paper, visit the original source: