The hype is not going to stay forever but I think the personal assistants like Google Home or Alexia will stay, virtual friends like Replika will stay, and also cost saving in service are going to stay despite some customers preferring humans.
Chatbots replacing apps didn't happen.
However, if your chatbot performs terribly bad in user tests, you may want to remove it from your platform. A good chatbot can do great, but a bad chatbot is bad. And it's not necessarily bad because of the technology, it's most likely simply a bad chatbot like many other chatbots. Making a good chatbot is not trivial and requires a lot of work and user testing.
I was working on a machine learning Open Domain Chatbot for a local museum. Our team tried to teach it its data so it could give info about it while also being able to have some small talk in addition to all the Information.
It worked fairly well but it will never be "very good" in my opinion since you will not be able to fully mimic a conversation with an actual human being.
So the future is kind of unsure but I doubt something big will happen with it in the near future.
I‘m not sure about ‚conversational‘ chatbots, they might work in a very narrow use-case only right now but using very task specific bots in e.g. Slack to automate tasks is definitely a great thing.
Chatbots replacing apps didn't happen.
However, if your chatbot performs terribly bad in user tests, you may want to remove it from your platform. A good chatbot can do great, but a bad chatbot is bad. And it's not necessarily bad because of the technology, it's most likely simply a bad chatbot like many other chatbots. Making a good chatbot is not trivial and requires a lot of work and user testing.