I started my career working on a large legacy project. It was the best learning experience, and super interesting. It certainly did not make me hate legacy projects. Work was like an investigation into who did this weird thing and when and for what reason - and there was usually a good reason. Then detangling if that reason is still valid today, or if we have time to refactor the thing appropriately to rewrite it in a different/better way, or how we can solve this obscure bug without breaking five other usages, etc. And doing all that as elegantly and non-disruptively as possible with hundreds of people on a tight timeline relying on the codebase. It was an extremely satisfying challenge.