Nurses Don't Hate AI—They Hate Being Surveilled

It starts with a beep. A notification that a patient's call light has been on for too long. A dashboard showing that Nurse Rodriguez spent 47 seconds less in Room 12 than the algorithm expected. An automated email to her supervisor, timestamped and triangulated, suggesting she might have missed a pain reassessment.

When I read through the comments on a Local News Matters piece about Kaiser nurses complaining about AI and surveillance, that's the picture that emerged. Not a Luddite rejection of technology, but a deep, visceral anger at being tracked, measured, and second-guessed by systems that seem designed to catch them doing something wrong.

But here's what struck me: buried in those 343 comments was a recurring thread of nurses saying some version of "I don't mind tech—I just want it to help me, not police me."

That's a different story. And it's one our data backs up.

The Real Problem Isn't AI

We track over 663 healthcare problems at PainSignal, and 416 of them already have app ideas attached. When you look at the ones that nurses are screaming about—the ones that consistently hit severity 5/5—they're not about AI. They're about unsafe staffing. Documentation that eats up half a shift. Medication errors that happen when you're covering twice as many patients as you should be.

ChartFlow Pro, an AI-powered charting assistant, is one of our most upvoted concepts. SafeStaff Nurse addresses the staffing ratios that make nurses want to quit. These aren't anti-tech ideas—they're pro-nurse ideas that happen to use AI.

The article painted AI as the villain, but nurses in the comments repeatedly distinguished between helpful automation and punitive surveillance. One nurse wrote, "I'd love an AI that drafts my notes and flags orders I might have missed. I hate the AI that tells my manager I took an extra three minutes on break."

That nuance gets lost in headlines.

What Nurses Actually Want

Looking at our problem data, three categories dominate the nurse experience:

Documentation burden. Charting takes hours. Nurses do it during lunch, after shifts, before shifts. One PainSignal submission described a nurse who comes in 30 minutes early every day, unpaid, just to review charts from the night before so she doesn't start her shift already drowning. An AI tool that drafts notes, pulls in relevant history, and flags inconsistencies isn't surveillance—it's a life raft.

Unsafe staffing. We see problem after problem where nurses describe patient loads that make them fear for their licenses. The article touched on this, and our data shows it's a top-three driver of burnout. Technology that predicts patient acuity and helps managers staff appropriately—instead of just tracking whether nurses are working fast enough—could literally save lives.

Fear of retaliation. SafetyReport Shield is a problem we track that speaks directly to something the article mentions: nurses being afraid to report unsafe conditions. If a surveillance system is watching your every move, will you speak up when you're asked to cut corners? Probably not. A truly useful tool here would provide anonymous, protected reporting channels—not more monitoring.

The Builder's Opportunity

The Hacker News thread got heated around this point: Are builders part of the problem, or part of the solution?

If you're a vibe coder looking at healthcare, you could very easily build something that makes nurses' lives worse. Slap a dashboard on some EHR data, add "AI insights" that are really just time-motion studies, and sell it to administrators who want to optimize their workforce. Congratulations, you've just built a surveillance tool.

Or, you could solve for the actual pain points. Our NurseFlow: Admin Reducer idea crowdsources ways to cut the administrative fat. Things like automated prior auths, shift-swap coordination, and medication reconciliation that don't require a nurse to click through seventeen screens.

The nurses in that thread aren't asking for less technology. They're asking for technology that respects their expertise and makes it possible for them to actually nurse.

One Thing to Keep in Mind

If you're building for this space, don't just parachute in. The Hacker News comments are full of nurses who've been burned by tech implementations that promised the moon and delivered a camera in the break room. The trust deficit is real.

But that's also an advantage for indie hackers who can move fast and iterate based on real feedback. Unlike the enterprise vendors that Kaiser might contract with, you can actually talk to nurses, understand their workflows, and build something they'll advocate for—instead of something imposed on them.

The 416 app ideas on PainSignal aren't theoretical. They're filled with specifics: "I need a tool that lets me document a wound assessment by talking, not typing." "I want a scheduling app that actually accounts for patient acuity, not just headcount." These are actionable, precise, and validated by the people who'd use them.

So if you're scrolling through Hacker News and getting depressed about the state of healthcare tech, look closer. The anger isn't at AI. It's at being treated like a machine. Build something that makes nurses feel like humans again, and you'll have a market that's begging for what you're selling.