The latest salvo from Google in the health specific LLM space.
Well, at least Google is being honest about its motivation: Greed. They just want to make money.
It also raises the issue of why the local system might actually allow this technology to be used instead of earlier efforts to achieve a broader goal which failed to proceed beyond local pilots despite those pilots being successful.
Lots of the medico-specific stuff being produced is probably going to die soon after implementation. Patients and doctors (rightfully) won’t trust LLMs to make diagnoses or attempt anything vaguely mathematical like risk stratification. It’s mostly the stuff that makes admin painless that will end up being useful. Begs the question of whether the problem is automating all of the admin, or whether all the admin is necessary in itself.
I think Google must still be in the testing phase rather than “roll out” as it doesn’t seem like they have been FDA approved yet. It will be interesting to see the results of the FDA process, CE marking, etc.