New insights from Google’s AI & Economy ATLAS

Google’s AI & Economy ATLAS reveals how artificial intelligence is rapidly transforming work across various sectors. The interactive, open-access experience allows users to explore the adoption rates of AI among diverse occupations, ranging from electricians to purchasing managers, as well as its usage at home and in different countries.

New research from Google, Google DeepMind, and MIT FutureTech sheds light on how scientists are incorporating AI into their work. The study, which examined 2,600 specialized AI models and surveyed over 600 scientists in the U.S. and U.K., found that scientists are leveraging AI at a higher rate than many other occupations, with nearly half of them utilizing AI daily.

While AI adoption among scientists is significant, the research indicates that scientists primarily utilize both specialized AI models and large language models (LLMs) like Gemini for different tasks. Specialized models are more prevalent in health and life sciences, while LLMs are widely employed across various scientific fields and task categories.

Scientists reported time savings of nearly seven hours a week thanks to AI, allowing them to focus more on research. However, the researchers also discovered that scientists spend considerable time validating AI outputs, face a backlog of untested hypotheses, and encounter bottlenecks in areas like physical experimentation and clinical validation.

These findings suggest that while AI holds immense potential to boost productivity, the ultimate impact on scientific discoveries may depend on the redesign of scientific processes and workflows to fully harness the capabilities of rapidly evolving AI technologies. As a long-term research project, ATLAS continues to work with academic partners to uncover new research areas and deliver fresh insights into how AI is reshaping the global economy.