"Clearly, OpenAI's advice not to use Whisper in high-risk domains, such as critical medical records, was a good one. But health care companies are constantly driven by a need to decrease costs by using seemingly 'good enough' AI tools"
What's terrifying isn't that the models hallucinated, it is that these companies are trusting it for these applications in the first place.
arstechnica.com/ai/2024/10/hospitals-adopt-error-prone-ai-transcription-tools-despite-warnings/
