1. Roomba.

2. Computers in an average modern car control steering, braking and acceleration to degrees not possible in the past.

It's clear that Kurzweil is more right than wrong if you read his predictions as trends and account for some degree of variation. Since he's talking about processes I am not sure what reading his predictions with a letter-of-the-law critical mindset buys you. The vast majority of what he talks about - has been invented - and is well on the way to becoming widespread. For me, it's the point of origin ("invention") of a technology and it's near-term projection that's most salient and Kurzweil nails it down.