This researcher is working on the "explainability problem" of #AI and outlines the urgent need for transparency.

If AIs (#NeuralNets, #MachineLearning, #LLM, etc) are increasingly used to make medical decisions, allocate loans or in policing, we don't want those decisions to come from nameless, faceless black box algorithms.

https://www.quantamagazine.org/cynthia-rudin-builds-ai-that-humans-can-understand-20230427/

oop looks like i failed to mention #phonology in my intro

i'm an #exemplar and #NeuralNets guy by disposition but i think there's real value in #formalist approaches (à la Bale & Reiss, or the #subregular stuff); they offer a distinct mode of understanding (in a vaguely #Marr way but i've come to believe the notions of "levels" has been detrimental overall)

some day i'll get it together enough to get a piece on #explanation and #pluralism of #theory written somewhere

#linguistics

#MichaelShermer - How #BeliefSystems Affect #Believers

How do belief systems work their credulous magic? What principles enable belief systems to influence individuals and commandeer groups?

https://www.youtube.com/watch?v=btAHRULz3l4&ab_channel=CloserToTruth

#Philosophy #Psychology #NeuralNets #NeuralNetworks #OriginsOfBelief #Tribalism #Rationalism #Rationality #Politics