MetaHealthResearch

Meta shares 10 health research manuscripts made by medical experts working with Muse models

Meta AI ·

Header art from the Meta AI research blog post: blue nodes joined by thin lines on a pale background

Meta built an internal coding scaffold on Muse Spark 1.3 and Muse Wonder with access to public data and a licensed database of about 5 million de-identified electronic health records. Over four weeks, a cohort of medical researchers used it to brainstorm hypotheses, write and run all the code and co-write manuscripts, while keeping authority over the scientific decisions and conclusions. Meta is sharing 10 resulting manuscripts in areas such as infectious disease, emergency medicine, oncology and sports medicine; each was reviewed by peers, a biostatistician and contracted outside experts, whose summary reviews are published with the papers.

How you’d use it: Health researchers can see how experts steered an AI coding scaffold at each research stage.

Sources

← All AI news

Get the daily AI news

The day's important AI news by email: short, plain summaries, each linked to its source. Unsubscribe any time.

Prefer email? Write to [email protected] ([email protected]).