Examining the use of artificial intelligence and large language models at TGen
Artificial intelligence (AI), the ability of machines to perform human-like functions, and large language models (LLM), a type of AI trained on pattern recognition in text data, are new technological architectures with potential applications across all divisions at TGen. However, the specific uses of these cutting-edge technologies within the institute remain unexplored. There is an unmet need for TGen to comprehend employee familiarity, usage, and interest and create trainings that leverage AI tools responsibly, benefiting both the institution and the patients we serve. To address this need, an institution-wide survey was devised and deployed, with the goal of understanding TGen employees’ current views of AI and LLM platforms. Using the survey data, the Science Writing Team developed TGen-centric presentations illustrating use cases for the two most prevalent AI types: text and image generators. Results from 133 total survey respondents revealed that although 40% of TGen employees have minimal to no familiarity with AI programs, 55% are excited about AI’s capabilities. The primary concerns around AI/LLM usage included incorrect information, referencing, and plagiarism. Additionally, 80% expressed confidence in TGen’s ability to establish appropriate AI guidelines, reflecting an overall positive attitude towards the implementation of these technologies. Practical applications tested for text generators include translating complex writing into outlines for philanthropic pitch pieces, social media blurbs, and sections of scientific publications. Moreover, image generators, while unable to seamlessly generate images with text, can create symbols useful in visual depictions of scientific concepts. These findings highlight the myriad potential benefits of integrating AI and LLM technologies into TGen’s operations.
