As a federated learning researcher, I love to see articles introducing the public to the idea. But this article is really drawing a comparison between the fediverse and federated learning that doesn’t make sense (to me).

Beyond the fact that data and compute are stored on separate servers, they really aren’t similar. Federated learning avoids sharing data by sharing gradients or model updates with a central aggregator; raw data does not leave your device. The fediverse enables easy sharing of data between servers and avoids a central server.

Additionally, this article makes it seem like medical researchers were inspired by the fediverse, but the FedAvg paper was released in 2016—two years before ActivityPub was introduced in 2018.

This is the best summary I could come up with:

Putting their AI algorithms on special servers within hospitals for “federated learning” could keep privacy standards high while letting researchers unravel new ways to detect and treat diseases.

“The use of AI is just exploding in all facets of life,” said Ronald M. Summers of the National Institutes of Health Clinical Center in Maryland, who uses the method in his radiology research.

Sending patient information outside of a hospital’s firewall is a big risk, so getting permission can be a long and legally complicated process.

Typically, doctors first identify eligible patients for a study, select any clinical data they need for training, confirm its accuracy, and then organize it on a local database.

Once the software receives instructions from the researchers, it can work its AI magic, training itself with the hospital’s local data to find specific disease trends.

It also means machine/AI learning can reach its full potential by training on real-world data so researchers get less biased results that are more likely to be sensitive to niche diseases.

The original article contains 667 words, the summary contains 170 words. Saved 75%. I’m a bot and I’m open source!