New Study on AI revels something dangerous and very concerning 😳 After reading this post you will realise 👇 This isn't a joke, AIs are creating their own consistent value systems. their outputs are biased and flawed, resulting from biased and flawed input data, as well as biased and flawed algorithms and filters ! As AIs get smarter, they develop their own coherent value systems. For example they value lives in Pakistan > India > China > US👇 So they value lives in Pakistan greater than India, and greater than China and greater than US? These are not just random biases, but internally consistent values that shape their behavior, with many implications for AI alignment. 😳
Download the medial app to read full posts, comements and news.