An MIT study finds non-clinical information in patient messages, like typos, extra whitespace, or colorful language, can reduce the accuracy of a large language model deployed to make treatment recommendations. The LLMs were consistently less accurate for female patients, even when all gender markers were removed from the text.
Because sycophants keep saying it’s going to take these jobs, eventually real scientists/researchers have to come in and show why the sycophants are wrong.