I was thinking this same thought. If the anthropic principle would be applying here (if the LHC does not run in futures containing human observers) then we are in a strange middle-land where -- actually -- the observance of a likely future shut-down is evidence against the anthropic principle.

Or is it?

If the LHC selects anthropic futures in which the LHC never runs, then this will occur through the selection of nodes that lead to the LHC never running. Thus, the current problems in the LHC could be such a node that would be selected. However, the selection would not occur until after the LHC could have run. So I'm convinced again that finding ourselves in a branch where the LHC won't work for the next two years is just evidence that the LHC was unlikely to over the next two years, anyway, independent of the anthropic principle.

Unless the LHC was supposed to run yesterday.