That gene used to be functional in lower mammals. The designer must have known it was there and worked at some point since he/she must have made it. Did it just naturally deteriorate in the new environment since there was no selective pressure? Or did the designer just throw in a broken gene for the hell of it?
You keep saying "the designer" as if you are assuming it is a single entity. Are you making the assumption that "the designer" is God? And then making assumptions about the attributes and qualities of "the designer" that would preclude errors?
In the software realm, applying a fix or an enhancement here can break something that was previously working there. And unless you have a very complete regression test environment (and most places don't even have a regression test environment at all), the software just gets released into the environment with the new feature and something broken. And of course, no one knows until the software users start complaining. Which, depending on the environmental conditions, will either be somewhere between right away to never (never in the case where what is broken is not used in its current environment.)