I think they're talking about conjectures that are unproven but seem "likely true" and people build further math off the assumption. E.G Reimann hypothesis