No abstract

Significant health disparities throughout the lifespan have long been evident within the American, non-White population. These differences include increased incidence of infant mortality, childhood hospitalization, diabetes, mental and behavioral health problems, cardiovascular disease, and late-stage cancer diagnoses. These types of disparities could be addressed by identifying those individuals at greatest risk for chronic disease or higher mortality rates and improving their health literacy and access to care. However, this is only possible if health management organizations and care providers have accurate racial and ethnic information. As an initial step at better understanding the problem, a survey of demographic data for Ohio Medicaid beneficiaries was conducted. The results of the study revealed inconsistencies and omissions in reported race and ethnicity for nearly 10% of records received from the state. This trend was most evident among older adults and those joining under the Affordable Care Act's Medicaid Expansion. Collectively, these results suggest that the first step in correcting health disparities may be to ensure that accurate information is available about the target population.

How do we gauge understanding? Tests of understanding, such as Turing's imitation game, are numerous; yet, attempts to achieve a state of understanding are not satisfactory assessments. Intelligent agents designed to pass one test of understanding often fall short of others. Rather than approaching understanding as a system state, in this paper, we argue that understanding is a process that changes over time and experience. The only window into the process is through the lens of natural language. Usefully, failures of understanding reveal breakdowns in the process. We propose a set of natural language-based probes that can be used to map the degree of understanding a human or intelligent system has achieved through combinations of successes and failures.

No abstract