A “ripple” that arrives too late or too soon, can have an impact the exact opposite of what is hoped.
Liset M de la Prida @LMPrida
Our latest paper with @acnavasolive & @adrubio1999 is out in @CommsBio A machine learning toolbox for the analysis of ripples across species https://nature.com/articles/s42003-024-05871-w In collaboration w @perpl_lab & @SAbbaspoor we found that AI algorithm trained in 🐁 could be applied to https://pic.twitter.com/MRVTgUOSkN
Replying to @LMPrida @acnavasolive and 4 others
Just being out of date, or without links and details for verification with reasonable time and effort – can change a safe source into a broken one. The AIs still are not open verifiable lossless or correctable by users — for all humans. Organizations, governments (organizations), industries (organizations), topic groups (organizations) all are making up and maintaining their own rules – mostly without feedback mechanisms of any sort. Let alone global open verifiable auditable ones with best in world, current methods.
A “ripple” that arrives too late or too soon, can have an impact the exact opposite of what is intended.
