I agree.

For the last months I have seen how, due a to a bad CI management, teams have started to add retries and larger timeouts in jest tests. The first causality was TDD: with four seconds loop is ok, but, failing first and waiting 30 seconds for that... nobody wants to do it.

But the second causality was quality. Now, this extra wait can make things work just for pure luck. Some internal coding or waits can be wrong, and race conditions arise.

So yes, auto-waits just seem an ugly idea.