I sometimes wonder what Kurzweil’s game is. Human-equivalent AI by 2029? (And as a commercial product?) The real-world plans to have self-driving cars available commercially by 2020 seem a bit ambitious - that just seems nutty. His prediction history is oddly inaccurate, despite his protestations, and I’m particularly baffled by his medical predictions. He’ll predict that some medical therapy will be in widespread use in 20 years when it doesn’t even exist currently in the lab; these are the sorts of novel treatments that would take 20 years to go from early clinical trials to commercial product, and surely he knows that. Frankly, I think he gives things too-early dates to generate controversy rather than because he actually believes it.
