I dont think anything will change until Dr Jayne and others take my approach of naming names, including how much…
EPtalk by Dr. Jayne 10/30/25
Sometimes reader comments make my day. One did earlier this week, when Data Diva accused hospital boards of “trying to cosplay as tech bros” and suggested that we consider “automating leadership bloat before automating bedside care.”
This inspired me to see whether a large language model could do a better job than hospital administrators at certain tasks. I asked Microsoft Copilot to assume that I am a hospital administrator trying to figure out how to make life better for the nurses who work at my hospital. I then asked, “How would they like to be rewarded for their hard work?”
I was pleased with Copilot’s response. Nurses deeply value recognition that feels personal, meaningful, and supportive of their well-being and professional growth. A mix of financial, emotional, and developmental rewards works best.
The response went on to suggest a combination of authentic recognition and appreciation, such as peer-to-peer recognition programs, along with fair compensation and financial incentives. Competitive salaries were specifically mentioned, along with ensuring that pay aligns with industry standards and reflects experience and performance. Performance-based bonuses and spot rewards were listed, as were tuition reimbursement, paid training, and clear opportunities for promotion and skill building.
Other suggestions included flexible scheduling, additional time off, and mental health support. Copilot went on to recommend that recognition be embedded into daily culture, “not just during Nurses Week,” and that rewards should be tailored to individual preferences to “ensure all staff feel seen and valued.”
It went on to ask if I wanted help designing a nurse recognition program tailored to my hospital’s culture and budget. I threw out a random number and asked what I could get for $100 per nurse. Suggestions included a customized thank-you box; a voucher for an experience, such as a massage or yoga class; branded gear, such as a high-quality fleece jacket, tumbler, or tote; a continuing education stipend; extra PTO, a coupon for a flexible shift swap; or a gift card for healthy meals.
Nowhere did it recommend pizza parties or challenge coins. For that alone, I can conclude that LLMs are better than actual hospital administrators. I ran these items past a couple of nurses and they were on board. Administrators should take note before they wind up being replaced by an AI assistant.
I was feeling a little punchy, so I went on to ask, “Do nurses like pizza parties?” Copilot was again accurate: pizza parties are appreciated as a kind gesture, but most nurses view them as insufficient on their own. They prefer meaningful recognition, support, and resources that address their real challenges. Copilot went on to suggest that pizza parties are “symbolic but shallow” and “can feel tone deaf” since they don’t address deeper needs such as burnout, staffing shortages, and lack of support.
Without prompting, it instead recommended authentic recognition, work-life balance elements such as flexible scheduling and adequate staffing, professional growth and career advancement opportunities, and mental health support. It went on to recommend that “if you still want to host a pizza party” that leadership should pair it with something meaningful and also make it inclusive and convenient, specifically recommending making sure that the night shift can participate. Winning the hearts of the night shift is pretty smart, so two points for Copilot.
I’ve had some medical adventures over the last year and have several important physician appointments pending. I’m always tuned in when I receive an email or text saying that I have a new message in my chart. I admit it triggers a bit of a fight-or-flight response. I was less than thrilled when I logged in to find that the message was letting me know that the hospital is having phone issues I should use the patient portal instead if I need to contact a physician. Health systems should be able to flag these kinds of communications as “non-urgent” or “a general communication from your health system” header so that patient anxiety isn’t provoked.
Pet peeve of the week: people who keep sending broken web links even though you’ve told them that the link is broken. I have been working with a vendor rep who keeps sending me documents to review. I dutifully report the broken links, but each subsequent includes the same broken links. If I can’t trust that you’re reading my emails and taking action on my requests to send content that I can actually view, I’m not sure you’re the kind of person or organization with whom I’d like to do business.
This article about AI-free periods for physicians caught my eye. It points out concerns for “deskilling” that is due to overreliance on technology. Singapore’s National University Health System has been implementing the AI-free periods after studies found that physicians who relied on AI tools during endoscopy were less able to use their own skills to detect polyps when the tool was taken away. By removing AI tools occasionally, leaders hope that physicians will maintain their core competencies and avoid being overly reliant on tools.
The article also mentions tracking physician performance to determine whether AI tools are having a negative impact. Anyone who has had to navigate a downtime situation when it hasn’t been practiced for a while knows what it feels like when technology is taken away, so I think that considering AI-free practice on occasion is a good idea. And if you haven’t had a downtime drill in a while, there’s no time like the present.
My hospital recently did a user survey to gather information on how well the informatics team did as they implemented a new feature. We’ve used anonymous surveys before and have always found them to be a good way to get direct user feedback and ideas for improvement. We don’t usually receive a lot of free-text user feedback, but we always get a few constructive comments.
This time, however, I was completely floored by how hostile some of the free-text responses were. Some of them even included personal attacks on members of the training and implementation teams.
I’ve seen enough anonymous posts on social media to know that societal rules are evolving to a point where people feel emboldened to say whatever they feel, but I haven’t seen these kinds of borderline threatening responses from our medical staff. It makes me wonder about the overall stress level of providers in the organization and whether we need to take additional steps to ensure the welfare of our employees.
Have you noticed a change in the level of civility at your institution? Have steps been taken to improve communications and ensure that staff members are safe? Leave a comment or email me.
Email Dr. Jayne.

I think it’s interesting that CoPilot does so well, likely because their training encourage truthful responses. These are probably not specific “nursing motivation” benchmarks, either, so there’s likely room for improvement. If we used a fraction of those scoring metrics for our management, we might see similar outcomes.
If we handed over management to the LLMs, though, it wouldn’t take long before the billionaires that run them, in collaboration with their business partners, decided to add “fiscal responsibility” to the training, and you’d find yourself in a reality where pizza parties are the best motivator (cheese-only, single slice max, during lunch break).