I think he just brought up the "int overflow into Double" thing so he could pimp out the double-dispatch type inference a little bit later:
> "... and so all this stuff about having a type-tag saying this is an int – which might not actually be technically correct, if you're going to overflow into a Double, right? Or maybe you're using an int but what you're really using is a byte's worth of it, you know. The runtime can actually figure things out around bounds that are undecidable at compile time."
I got the distinct impression that he wasn't arguing against type inference, merely said that the classic way to do it doesn't always work with dynamic languages, and there are better (newer) ideas regarding how to infer types without resorting to type tags.
> "... and so all this stuff about having a type-tag saying this is an int – which might not actually be technically correct, if you're going to overflow into a Double, right? Or maybe you're using an int but what you're really using is a byte's worth of it, you know. The runtime can actually figure things out around bounds that are undecidable at compile time."
I got the distinct impression that he wasn't arguing against type inference, merely said that the classic way to do it doesn't always work with dynamic languages, and there are better (newer) ideas regarding how to infer types without resorting to type tags.