My ears perked up at the dual statement that although you can write a Lisp program that writes Lisp programs, nobody does it (um, macros?); and that you can't write a C program that writes a C programs (okay, maybe nobody does this, but you certainly can). I guess he means that C is not "aware" of its own constructs the way that Lisp is.

Also, he said that the algebraic languages were "dying out", but that because of this self-generating ability (which nobody uses), that the future of Lisp was "open".

Huh?