Tim Hwang argues AI models lean toward Buddhist viewpoints due to enterprise safety and alignment training · Digg