Marcel D comments on My attempt at explaining the case for AI risk in a straightforward way