The AI Singularity

Much is being made of the advances in machine learning and AI in general. So much so that many tech people are convinced that human-level intelligence is less than a decade away, and that soon we'd have nothing to do (because all research, work, everything, will be done by the AI).

Is such a future possible? That too in less than 10 years from now?

It's interesting to note that the people proclaiming the rise of a human-level-intelligent AI are not practitioners from this field. They are entrepreneurs, heads of tech department, physicists, and suchlike. What gives these people the confidence to predict these things is beyond me. And what is even more beyond me is that nobody seems to care to ask the actual researchers what's going on.

I don't want to spend too much time on this frivolous issue, so I'll present two links that will show you just how horribly unintelligent and blind are today's ML systems:

So, for the umpteenth time, folks, please calm down and use your heads. It's in the selfish interests of these individuals to make such proclamations – these either keep their companies popular (and this preserver “value” in the eyes of the investors) or serve to bring to limelight an otherwise forgotten celebrity.

Keep calm and carry on!