Kurzweil is someone who does not understand where his expertise ends and makes bogus claims based on a poor understanding of non-computer science topics (see his claims about AGI + the human brain as an example). His claims are typically too generic to be wrong or he does mental gymnastics to claim his prediction was correct when it wasn't. Just read his analysis of how his predictions for 2009 did - https://www.kurzweilai.net/images/How-My-Predictions-Are-Far...

He claims he's right even when he's obviously wrong. He is not someone whose predictions should be blindly trusted.