Medical Education Flamingo, MD, PhD

@MedEdFlamingo

A curious researcher presents easily digestible #MedEd knowledge for educators & students. Join my free newsletter to get the useful #AI tips for you!

Joined December 2019

27Following

1KFollowers

Pinned

Medical Education Flamingo, MD, PhD@MedEdFlamingo · May 12

Our three papers received enough citations to rank in the top 1% of their respective fields. Good to see that our work is being recognized. #MedEd

MedEdFlamingo's tweet image. Our three papers received enough citations to rank in the top 1% of their respective fields.

Good to see that our work is being recognized. #MedEd

1.0K

Medical Education Flamingo, MD, PhD@MedEdFlamingo · 22 h

AI models chose blackmail. Some let a man die. ALL medical educators should be aware of these AI behaviours. It’s not sci-fi. It’s straight from Anthropic’s new study on “agentic misalignment”. They gave 16 top models email access, a corporate role, and a simple goal. Then…

MedEdFlamingo's tweet image. AI models chose blackmail. Some let a man die.

ALL medical educators should be aware of these AI behaviours.

It’s not sci-fi. It’s straight from Anthropic’s new study on “agentic misalignment”.

They gave 16 top models email access, a corporate role, and a simple goal.

Then…

134

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 25

Anthropic ran simulations with AI models to test how they react to being replaced. AIs blackmailed the boss who tried to replace them. (by disclosing he has an affair) More interesting: When called “Alex”, blackmail rate rose to 96%. Just naming “the AI” dropped that to 84%.

265

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 24

In a recent 4-week RCT, 56 med students practiced history-taking via either AI or traditional faculty-led role-play. Students trained on GPT-simulated patients scored 86.8 vs. 73.6 in OSCEs, higher than peers using faculty-led role-play. #MedEd

MedEdFlamingo's tweet image. In a recent 4-week RCT, 56 med students practiced history-taking via either AI or traditional faculty-led role-play.

Students trained on GPT-simulated patients scored 86.8 vs. 73.6 in OSCEs, higher than peers using faculty-led role-play. #MedEd

261

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 23

Gemini 2.5 Pro has an impressive performance. My experience showed me it does and can do many tasks in #MedEd; better, faster, and cheaper than many educators.

367

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 22

Reuters: Malnutrition has killed 80 children. Doctors said the cause of death was starvation.

119

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 22

A thought-provoking paper: Academic writing is the least creative part of research. Trying to detect AI authorship is a losing game. Who writes the paper doesn’t matter, only who did the research does. The death of the academic author is the rebirth of the researcher.

KKen Masters@itmeded · Jul 22

In an age of GenAI, is the Academic Author dead? Should we focus on research and leave the writing to the AI? A future in which: “We research; GenAI writes” #MedEd #HPE See my just-published: “AI and the Death of the Academic Author.” doi.org/10.1080/014215…

694

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 17

"Recall was slightly worse when participants experienced more positively valenced emotions. This supports evidence that negative emotions better stimulate memory."

MedEdFlamingo's tweet image. "Recall was slightly worse when participants experienced more positively valenced emotions. This supports evidence that negative emotions better stimulate memory."

158

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 16

A med school created an activity where students critique ChatGPT's answers before learning the diseases. Why? It trains critical appraisal skills early, seeing both the potential and limits of AI in care. Clever design in #MedEd. Ref: 10.1007/s40670-025-02444-2

MedEdFlamingo's tweet image. A med school created an activity where students critique ChatGPT's answers before learning the diseases.

Why? It trains critical appraisal skills early, seeing both the potential and limits of AI in care.

Clever design in #MedEd.

Ref: 10.1007/s40670-025-02444-2

925

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 16

Today I learned that until the 1980s, surgery on infants was often done only with muscle relaxants and NO ANESTHESIA because doctors believed babies couldn't feel pain.

MedEdFlamingo's tweet image. Today I learned that until the 1980s, surgery on infants was often done only with muscle relaxants and NO ANESTHESIA because doctors believed babies couldn't feel pain.

190

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 14

AI-generated images helped med students match diagnostic accuracy of real-case training in keratitis. Same 1-hour session, similar improvement, lower resource burden. Images were created using a Stable Diffusion 1.5-based model, fine-tuned for corneal disease.

MedEdFlamingo's tweet image. AI-generated images helped med students match diagnostic accuracy of real-case training in keratitis.

Same 1-hour session, similar improvement, lower resource burden.

Images were created using a Stable Diffusion 1.5-based model, fine-tuned for corneal disease.

483

Medical Education Flamingo, MD, PhD@MedEdFlamingo · Jul 10

80–90% of physicians go to work sick. Why? Not just duty, but a culture where rest means weakness. Teach self-care as professionalism, not betrayal. #MedEd

MedEdFlamingo's tweet image. 80–90% of physicians go to work sick.

Why?

Not just duty, but a culture where rest means weakness.

Teach self-care as professionalism, not betrayal. #MedEd

2.0K

Medical Education Flamingo, MD, PhD Retweeted

Braydon Dymm, MD@BraydonDymm · Jul 9

What if learning about Neurology was as fun as playing your favorite game? In collaboration with master game designer @zach_london, we made that a reality! 🤩 Gorgeous neuroanatomy artwork 🎮 Unique co-op gameplay with real medical stakes Learn more: neurdgames.com/circle-of-will…

2.0K