Shane Mueller worked with other collaborators in an interdisciplinary effort to conduct and disseminate research on the need and assessment of efficient AI. The development of AI often takes many stages of development and requires significant investments of effort and money, both of which can be seen as hinderances in the return margins of AI. Dr. Mueller and his collaborators developed an assessment of AI efficiency with interdependent human systems to determine if the work system is learnable and useful, which can be expanded upon to conclude whether or not they achieved the desired outcomes of investments.
The work is entitled “‘Minimum Necessary Rigor’ in empirically evaluating human-AI work systems.”
Human-AI (HAI) systems are built on human-AI collaboration, as the name implies. Humans utilizing these systems work interdependently with AI, utilizing AI tools to complete work tasks. These systems are present in many important sectors, such as “emergency response management, industrial process control, health car, business, banking and finance, military command and control, autonomous vehicles, transportation systems, weather forecasting” and, well, you get the idea. The research done by Dr. Mueller and others will aid the efforts of developing and refining HAI systems that provide a strong return on investments of time and money.