IntelliJ resolves the functions just fine, so why is gradle giving this error?

x is Float

does not imply

Float : T

, which would be required for that

return

.

This is counterintuitive. T is inferred as Float and Float is Number, but compiler just says excuse me I need T why you give me Float? I mean for non-final classes it's true, but for primitives and final classes, it's weird.