Self driving cars? Not so fast.

I suspect this article is, well, not exactly unbiased, but the fact remains clear as day that driverless technology is proving to be a very, very much harder problem than initially anticipated. And, in fact, going all the way back to Marvin Minsky in the 1970s, the predictions of artificial humanlike mental capacities and capabilities have invariably been overestimated. Driverless cars need not be self-aware, but it took evolution over a billion years to create a prototype large animal that could avoid getting itself killed long enough to reproduce (enough of the time), so it's perhaps a bit arrogant to think we can achieve the equivalent in a decade or two. I don't doubt it's possible, but it's not going to be easy or quick. And as for "AGI," artificial general intelligence, which it's usually just presumed will be somehow self-aware, technology is nowhere near. Minsky thought we'd have it by 1980. But since current AI methods don't even address the question of functionality of consciousness, and current AIs have not just a little but quite literally zero self-awareness, I think the prospects for "humanlike" intelligence (including self-awareness and intention) anytime in the foreseeable future can be discounted all the way down to "effectively zero."