In a conversation at this year’s rich person convention—aka the World Economic Forum—Microsoft CEO Satya Nadella warned that AI will lose public support unless it’s used to “do something useful that changes the outcomes of people and communities and countries and industries.”
He did at least provide one real example of what he means by all this: “When a doctor can … spend more time with the patient, because the AI is doing the transcription and entering the records in the EMR system, entering the right billing code so that the healthcare industry is better served across the payer, the provider, and the patient, ultimately—that’s an outcome that I think all of us can benefit from.”
Even the transcription software that is not purely transcription, but actually an LLM that “takes notes” and “summarizes” can hallucinate and add in stuff that was never mentioned. And then AI trained on biased data (i.e., like all of them) tend to give shittier notes for any patients who are of color.
Oh definitely. That’s exactly why I specified that it needs to be explicitly transcription software. Even speech to text on my phone gets it wrong with enough regularity that I check it every time. I can’t imagine what it would be like if I was using less common words In a medical setting. I don’t love the idea of every word said in a doctor’s office being recorded and that recording being on record forever, but to a certain extent I can understand doctors who might think that would be helpful. What I don’t think would be helpful is having anyone except the doctor or another trained medical professional summarize that information. I do understand that doctors are human and mess up and miss things and might even take worse notes than AI would, but at least it is a Doctor who is doing that. It is a human being who met the other human being and sat in a room with them who is making these decisions.