Comment by socalgal2

I look at it like this. I read the book when he made the predictions. Many of them came true. So those predictions were useful to me.

It sounds like for you, the fact that they weren't perfect made them worthless and he's someone who's opinions you should ignore.

I see enough of his predictions as close enough that I think he was worth listening to. He predicted computers would equal the human brain by 2029. The majority of AI researchers now believe it will be even sooner. When he made the prediction, most people thought he was far far far off or maybe even talking science fiction. Yet here we are.

His biggest point was expoential progress very common in many domains. Whatever the state of something is, if it's improving expoentially it will seem like it's not close to being there yet, a short time before it's actually there. A lake that takes 30 days to fill and gets filled double the previous amount every day seems empty for 28 days, at day 28 it's 1/4th full so still looks mostly empty. At day 29 it's 1/2 full. I personally see is exponetial preductions as mostly true.

I think that will be true of self driving cars. They are later than predicted but they exist and are expanding quickly. Waymo is in 14 cities. How many times do they and their competetors have to double before they're ubiquitous? That idea was introduced to me by Kurzweil.

Others have pointed out, once robots can build things they'll be used to build robot factories and they'll explode in proliferation. I don't think Kurzweil made that prediction but I think his books and presentations on exponential growth influenced others to see it in more things.