The need for continuous sync arises when there is a requirement to sync project code and run commands remotely with ssh. Continuous sync allows for real-time updates and propagation of changes to the code on the cluster. It can be useful in workflows like a jupyter + nbdev setup, where changes need to be propagated to code in other notebooks/files on the cluster. The use of sky exec instead of manually running commands with ssh can also be considered as an alternative workflow.