We have absolutely no way to tell how far from "AGI" we are. What we know for sure is that we're not there yet. And what seems likely is that we're getting clo

We have absolutely no way to tell how far from "AGI" we are. What we know for sure is that we're not there yet. And what seems likely is that we're getting closer, and that's something. That is as much prediction we can get. I don't think that Compositionality is a wall, it is clearly an interesting feature, but I think that it is pretty clear by now that the Turing test or anything in the same spirit is far from sufficient.