I agree with the general skepticism regarding AGI, but I think it's worth pointing out that there is some progress towards robust representations of abstract co

I agree with the general skepticism regarding AGI, but I think it's worth pointing out that there is some progress towards robust representations of abstract conceptual knowledge (http://science.sciencemag.org/content/331/6022/1279 http://science.sciencemag.org/content/331/6022/1279) that are far beyond feature extraction.