RESUMEN
The study provides a comprehensive review of OpenAI's Generative Pre-trained Transformer 4 (GPT-4) technical report, with an emphasis on applications in high-risk settings like healthcare. A diverse team, including experts in artificial intelligence (AI), natural language processing, public health, law, policy, social science, healthcare research, and bioethics, analyzed the report against established peer review guidelines. The GPT-4 report shows a significant commitment to transparent AI research, particularly in creating a systems card for risk assessment and mitigation. However, it reveals limitations such as restricted access to training data, inadequate confidence and uncertainty estimations, and concerns over privacy and intellectual property rights. Key strengths identified include the considerable time and economic investment in transparent AI research and the creation of a comprehensive systems card. On the other hand, the lack of clarity in training processes and data raises concerns about encoded biases and interests in GPT-4. The report also lacks confidence and uncertainty estimations, crucial in high-risk areas like healthcare, and fails to address potential privacy and intellectual property issues. Furthermore, this study emphasizes the need for diverse, global involvement in developing and evaluating large language models (LLMs) to ensure broad societal benefits and mitigate risks. The paper presents recommendations such as improving data transparency, developing accountability frameworks, establishing confidence standards for LLM outputs in high-risk settings, and enhancing industry research review processes. It concludes that while GPT-4's report is a step towards open discussions on LLMs, more extensive interdisciplinary reviews are essential for addressing bias, harm, and risk concerns, especially in high-risk domains. The review aims to expand the understanding of LLMs in general and highlights the need for new reflection forms on how LLMs are reviewed, the data required for effective evaluation, and addressing critical issues like bias and risk.
RESUMEN
BACKGROUND: Asthma exacerbations cause lung hyperinflation, elevation in load to inspiratory muscles, and decreased breathing capacity that, in severe cases, may lead to inspiratory muscle fatigue and respiratory failure. Hyperinflation has been attributed to a passive mechanical origin; a respiratory system time-constant too long for full exhalation. However, because the increase in volume is also concurrent with activation of inspiratory muscles during exhalation it is unclear whether hyperinflation in broncho-constriction is a passive phenomenon or is actively controlled to avoid airway closure. METHODS: Using CT scanning, we measured the distensibility of individual segmental airways relative to that of their surrounding parenchyma in seven subjects with asthma and nine healthy controls. With this data we tested whether the elevation of lung volume measured after methacholine (MCh) provocation was associated with airway narrowing, or to the volume required to preventing airway closure. We also tested whether the reduction in FVC post-MCh could be attributed to gas trapped behind closed segmental airways. FINDINGS: The changes in lung volume by MCh in subjects with and without asthma were inversely associated with their reduction in average airway lumen. This finding would be inconsistent with hyperinflation by passive elevation of airway resistance. In contrast, the change in volume of each subject was associated with the lung volume estimated to cause the closure of the least stable segmental airway of his/her lungs. In addition, the measured drop in FVC post MCh was associated with the estimated volume of gas trapped behind closed segmental airways at RV. CONCLUSIONS: Our data supports the concept that hyperinflation caused by MCh-induced bronchoconstriction is the result of an actively controlled process where parenchymal distending forces on airways are increased to counteract their closure. To our knowledge, this is the first imaging-based study that associates inter-subject differences in whole lung behavior with the interdependence between individual airways and their surrounding parenchyma.