On Andrew Gelman’s blog, “It’s kinda like phrenology but worse.” It discusses an ML paper that supposedly learns to predict “trustworthiness” from images of faces. No, actually, not images of faces, but portraits of faces over the last few centuries. Balderdash, the whole thing.
On Andrew Gelman’s blog, “It’s kinda like phrenology but worse.” It discusses an ML paper that supposedly learns to predict “trustworthiness” from images of faces. No, actually, not images of faces, but portraits of faces over the last few centuries. Balderdash, the whole thing.