There is always responsibility with Railway, that's given. But also taking into account how many big websites went down when AWS was down, building critical redundancy at such large scale is not cheap, and not many companies do it. Same as security theatre, we have redundancy theatre because they needed to sell the CLOUD.