This is sort of what's wrong with HN lately. There seems to be a sweeping majority of Singularity believers that just take "AGI in 30-50 years" as just an obvio

This is sort of what's wrong with HN lately. There seems to be a sweeping majority of Singularity believers that just take "AGI in 30-50 years" as just an obvious inevitability. So therefore we just go straight into SciFi mode and make all kinds of speculation with that premise. Then, this couples with this sort of, I don't know how to explain it, maybe its like an uncritically overly anti-establishment or "naively disruptive" current that just collectively nods like "Oh yeah, the traditional law firm is dead meat." It's surprising to m,e as I believe the community is largely made up of technologists, that I never see much questioning the fundamental claims here. For that wet dream of that Internet Bayes Theorem Religion, that I will not name, to come true, it needs to first be possible and second, there must br credible, serious people to actually be working on it. Granted, artificial intelligence is a very hot area of research. However, for anyone that is either a researcher in this field, a graduate student in the field, or just someone actually very current on the latest papers and the current research programs, knows this is AGI of the SciFi sort is not really what is being worked on. Not to go into all of the field but sure there are many interesting things happening in pattern recognition, NLP, machine learning and data mining and much work done with things like convolusional neural networks ..... but I'm sorry to say that these an AGI Overlord do not make. They are fancy techniques that yield very powerful and interesting results but in consideration of an AGI they are rather boring and banal. Largely, the techniques have been discontinuous incremental improvements on reasonably simple ideas from the 1950s and 1960 but now utilizing massive parallelism. Unfortunately, I think, for AGI a non-trivial number of techniques are rooted in statistical and probabilistic methods, which I think anyone would agree is probably a a wrong track toward intuitive and creative problem solving. It's not that those of us critical of AGI proclamations think that it is impossible, that would be rather bold, or that we think humans are magical or something. We probably just think that our brains, or even abstractly "true intelligence", are things that are unlikely able to be simulated by something like a Turing Machine. Any skeptic knows that extraordinary claims require extraordinary evidence. And I gotta tell ya, the way facial recognition works or even Watson, ain't it. Also, I will say, just because something in the world is a certain way or if it is not exactly like Uber, does not mean it is necessarily a bad thing .... I'm just sayin' .... An AGI Overlord would probably not be like the wild-eyed disruptor but more contemplative and understanding of the world as being very complex.