Nvidia Announces AI-Powered "Agents" to Replace Nurses in Hospitals

Because we live in a dystopian healthcare hell, AI chip manufacturer Nvidia has announced a partnership with an AI venture called Hippocratic AI to replace nurses with freaky AI "agents."

These phony nursing robots cost hospitals and other health providers $9 an hour, a fee that barely falls above the US minimum hourly wage, and far below the average hourly wage for registered nurses (RNs.)

In a press release, Hippocratic AI described the disturbingly cheap nurses as part of an effort to mitigate staffing issues. The company also claims that the agents won't be doing any diagnostic work, and will instead be doing "low-risk," "patient-facing" tasks that can take place via video call.

Nvidia showcased what such a call might look like in a demo video in which a very inhuman-feeling AI "nurse" checks in with a post-appendectomy patient, offering aftercare advice and answering questions about whether certain antibiotics are safe for the patient, who says they're allergic to penicillin and diabetic, to take. The exchange goes swimmingly, as controlled, advertised demos generally do. Hippocratic AI also claims that according to its own research, AI agents generally outperform their human counterparts in most categories.

But this all said, is giving a hospital patient medical directives ever really a "low-risk" order? It's hard to believe that bottling pre- and post-care nursing tasks into what seems like a glorified customer service chatbot is the sound, ethical answer to the nursing crisis that Hippocratic AI lauds its tech to be.

What's more, public health experts and RNs themselves say the nurses' staffing crisis — which has culminated in historically large nursing strikes over the past few years — isn't the result of sky-high nursing salaries begging for a quick automation fix, but has instead been caused by hospital cost-cutting leading to abysmal working conditions, unfair contracts, and low wages.

"Labor is the main expense in healthcare, so how do you make money? You squeeze your labor costs," Janette Dill, professor of health policy and management at the University of Minnesota, told CNN last January amid large nursing strikes in New York City. "Your nursing workforce is your biggest workforce."

Hippocratic AI's marketing for its unreal nurses also strongly appears to misrepresent the average nurse's salary. On its website, the AI venture claims that based on survey data from 2017, the average "all-in" cost for a "permanent, full-time" RN is about $90 per hour — a number that, before taxes, would mean a nurse were raking in over $180,000 a year. (You know, because RNs famously make six figures.)

But as Futurism sister site Popular Science points out, a far more recent US Bureau of Labor Statistics survey from 2022 puts the average RN's salary at $38.74 an hour, which would come out to less than $80,000 — a full hundred grand or so less than Hippocratic AI's estimate! — a year.

"With Generative AI, the incremental cost of healthcare access and interventions is trending to zero," Hippocratic AI's website adds, arguing that language models "are the only scalable way" to close the gap between healthcare staffing demands and a lacking "workforce supply." Sure.

Look, hospitals are having staffing shortages, and those shortages do impact patient care. But the notion that six-figure nursing salaries are to blame — as opposed to a generally broken healthcare system — is deeply flawed. And broadly speaking, when it comes to nakedly job-replacing AI "agents," we'd all be better off just to call them as they are: a cheap way to automate labor that a company or institution already didn't want to pay for.

More on AI and healthcare: Doctors Using AI to Automatically Generate Clinical Notes on Patients