LLM's
Every article we have published: research summaries, product thinking and developments in health AI.
LLM's · 363 articles

AI Patient Simulators Train Doctors on Biased Stereotypes
Medical schools are using AI to train future doctors, but a new audit reveals these digital patients are quietly reinforcing dangerous demographic...

AI swarms overestimate ICU mortality risk
Letting AI models revise their own clinical decisions makes them more pessimistic and biased, not more accurate.

AI spots sepsis early using less data
Hospitals are drowning in false sepsis alerts, but a new machine-learning model proves we can detect this killer faster by cleaning up...

New AI workflow slashes clinical data costs
A new agentic framework extracts structured clinical insights from messy medical notes at a fraction of the cost of traditional LLM runs.

Hospital AI Tool Hallucinates Twelve Percent of Time
A new study reveals that while doctors find EHR-integrated generative AI easy to use, the software regularly invents clinical details and ignores...

Cheap AI matches flagship models on patient feedback
Hospitals paying premium prices for top-tier AI models to analyze patient feedback are throwing money away.

Why NHS AI Scribes Need National Oversight
Rushing AI dictation tools into clinics without centralized guardrails shifts the entire burden of safety onto overworked doctors and vulnerable patients.

AI models fail at medical citations
Even the most advanced artificial intelligence models frequently cite the wrong medical papers, but a new verification method might fix the problem.

Medical AI fails at real-world clinical reasoning
A new benchmark reveals that medical artificial intelligence aces multiple-choice tests but stumbles when faced with real patient cases.

AI scribes shift malpractice risk to doctors
As hospitals rush to adopt ambient AI scribes, clinicians are quietly signing off on a massive liability shift.

AI Search Models Target the Clinical Workflow
Medical AI is moving past generic chatbots to deliver specialized search tools that clinicians actually trust.

MISP-Bench: Decomposing User-Provided False Priors into Answer, Rationale, and Guard Effects
A new benchmark reveals that prompting AI to double-check its reasoning often backfires, forcing the system to agree with a patient's self-diagnosis.
Also on LinkedIn
AI Health Hub posts the same research summaries on LinkedIn every day. More than 3,000 people follow it there.