Source: symmetry (joint Fermilab / SLAC publication)

Authors: Laura Dattaro

Published Date: 2024-05-06T08:00:00-0500

Summary: In 2017, Savannah Thais attended the NeurIPS machine-learning conference and learned about bias in machine-learning algorithms from a talk by AI researcher Kate Crawford. A study by Joy Adowaa Buolamwini revealed that facial-recognition technology had significant gender and racial biases, misclassifying women of color 32% more often than white men. This exposure transformed Thais’ perspective, leading her to focus on the ethical implications of AI. Algorithmic bias is a concern in physics research, as machine-learning models can inherit biases from their training datasets. Physicists, like cosmologist Brian Nord, are working to push machine-learning capabilities while being mindful of potential pitfalls. Ethical considerations in AI are crucial, as the technology developed for scientific purposes can have broader societal impacts. Thais emphasizes the need for physicists to incorporate ethical frameworks into their work to improve machine learning’s application and to educate others on its implications.

Read full article here

By admin