Inefficient hello world programs (idea)

Many months ago, one listless Sunday afternoon, I crafted a stupidly inefficient program using the already long-winded assembly language. This is quite possibly the dumbest way to print "Hello world!" to the DOS standard output using NASM...

segment code

..start         mov ax,0B800h
                mov es,ax
                xor di,di
                mov cx,2000
                mov ah,7
                mov al,0
                rep stosw      ;Clear the screen (just to be safe!)

                mov dh,0
                mov dl,0
                xor bh,bh
                mov ah,2
                int 10h        ;Reset cursor to 0,0

                mov ax,data
                mov ds,ax

                mov ax,0B800h
                mov es,ax
                xor di,di
                mov bx,0

findchar        mov cl,0       ;Search every ASCII character from 0...
                cmp byte cl,[ds:_message+bx]
                je print
                inc cl
                jmp findchar   ;Until required letter is found

print           mov ah,7
                mov byte al,cl
                stosw          ;Get char and store character in stdout...

                cmp bx,12      ;Check for end of string...
                je end
                inc bx
                jmp findchar   ;Then do it all over again!

end             mov ax,4C00h
                int 21h

segment data

_message db "Hello world!"

A similar effect could probably be achieved in less than ten lines, although saying that, the above program would probably still be more efficient than cout << "Hello world!" (at least with my POS compiler)