I am writing a book and am seeking feedback on different chapters before submitting for final publication. Can I use the draft chapters as conference papers or at visiting lecture seminars?
That is up to the book publisher. Some will be fine with it, others not. Copyright might be one (blocking) issue. If you have given up copyright to some conference, then some publishers will reject your proposal, since they want to hold it and it is no longer yours to give. Lectures and such are no real issue, since you retain copyright.
Another thing that might influence a publisher (around the copyright issue) is how different the book chapters are from the earlier material. But even then, you have to be careful about self-plagiarism since "repeating yourself" in one paper of something in another needs to be handled delicately and with full disclosure.
Just make sure you retain rights (possibly a license) to anything that you publish and want to include in a book later.
You can, of course, publish many things on a topic in conference papers that don't conflict with any rights that you might need in order to publish a book.
There are a few different questions here. Are you "allowed" to talk about the work at conferences? I'd say "of course, it's not a secret". Possibly using your literal book chapter(s) as overheads would be inappropriate.
In any case, the point in the previous paragraph is that you are not prohibited from "ever mentioning again" your own prior work, or to-appear work.
What is possibly problematical is "getting credit twice" for one piece of work. And, relatedly, but differently, copyright violations. The "getting credit twice" prohibition in recent years (decades?) has struck me as strange, but, well, ... Yes, I understand that there are "popular" algorithms to (supposedly) compute one's "impact" or ... whatever... And writing more than one paper about a given idea is somehow cheating. I have no idea.
So, yes, on one hand, it is certainly reasonable to talk about your own work as much as you want... in my opinion. Perhaps to use any repeated element as a novelty-scoring mechanism is not honest. And publishers will have varying opinions about prior appearances of the literal text (not ideas) of what you give them.