Meta says AI scaffold helped medical experts write 10 health research manuscripts
Overview
Meta says a cohort of medical experts used a human-led AI research scaffold, built on Muse Spark 1.3 and Muse Wonder, to produce 10 health research manuscripts covering cardiology, oncology, infectious disease and emergency medicine.
The scaffold draws on 5 million de-identified electronic health records and public data. Meta says the experts used it for original research, hypothesis generation, code execution and manuscript drafting, while keeping full authority over scientific decisions.
Meta, through a post from Jason Wei, says the group wrote the ten manuscripts in about two weeks, and that researcher Paddy Ssentongo generated three of them. Alexandr Wang puts the timeframe at under four weeks. Meta's statements are the only source for these claims, and the manuscripts themselves are shared by Meta.
Written by AI from the articles below · updated Oct 14, 4:49 PM ET
Check the sources:
Developments
2 developments
- Oct 14, 2:52 PM ET · 1 articleMeta's health workstream tests AI scaffold producing ten medical research manuscripts with expertsJason Wei: Meta's AI tool helps medical experts write ten health research manuscripts in two weeks
- Oct 14, 2:44 PM ET · 2 articlesMeta shares 10 health research manuscripts produced with a human-led AI research scaffoldAI at Meta: Meta shares 10 health research manuscripts built with Muse Spark 1.3
Article timeline
The articles in this story. Times are ET.
Alexandr Wang@alexandr_wangOfficialMeta publishes 10 health science papers written with Muse modelsAIAlexandr Wang says 10 new health science papers were written with Muse models in under four weeks. Meta says a cohort of medical experts used a human-led research scaffold powered by Muse Spark 1.3 and Muse Wonder, with access to 5M de-identified electronic health records, to produce the manuscripts. The papers span cardiology, oncology, infectious disease, and emergency medicine.
Jason Wei@_jasonweiXMeta's AI tool helps medical experts write ten health research manuscripts in two weeksAIMeta says a scaffold built on Muse Spark 1.3 and Muse Wonder, with access to 5M de-identified electronic health records, helped a small group of medical experts write ten research manuscripts in about two weeks. Researcher Paddy Ssentongo generated three of them, and the experts said the setup let them test hypotheses faster. Meta says human researchers stay in charge of scientific decisions, and the manuscripts span cardiology, oncology, infectious disease, and emergency medicine.
AI at Meta@AIatMetaOfficialMeta shares 10 health research manuscripts built with Muse Spark 1.3AIMeta says it built a human-led research scaffold powered by Muse Spark 1.3 and Muse Wonder, using 5M de-identified electronic health records and public data. A cohort of medical experts used the system to produce original research and draft manuscripts while keeping full authority over scientific decisions. Meta is sharing 10 resulting manuscripts spanning cardiology, oncology, infectious disease, and emergency medicine.

Heat trend
- Current comparable heat
- 19
- Peak comparable heat
- 20Oct 14, 3:00 PM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.