This is an abstract from a talk at Cornell University...

[Norman Hardy]

I recently saw a video tape of a talk by Feynman on quantum computers. It
was taped in '86 or '89 I think. It was his presentation of 'balistic'
quantum computers. In that talk he refered very briefly to the David Deutch
idea of the many worlds computer but was sceptical, but not entirely
dismissive of it.

In other comments Feynman seemed to think that the many worlds idea was not
very useful. It would certainly be useful if it helped design such a
computer. I would bet against it.