Part of being senior is having enough business context to know when to pull late nights and weekends. If you're doing it because the gantt chart some PM created 18 months ago said project X would be done by next month but everything is off the rails, yes, it's foolish and unfair to work nights and weekends to make it happen. If you're doing it because re:Invent is in a month and your team has a new AWS service to announce, that's a much better reason. Or in the startup world, if there's a major client demo next week and you're quickly hacking in a new feature they've been asking for.