Information architecture happens to be one of my favorite things to discuss and teach to other people, so I have never allowed the thought of it being "dead" to enter my mind. What I can add here is that the work of an IA is getting even more challenging in this hyper complex world we are creating.
Most recently I have helped a company who acquired a few other companies to reassess their language and structures to align as one coherent thing. What it taught me is that IA work today cannot be done "for a client" -- but rather must be done "with a client" - and that change simply cannot happen overnight. It is incremental changes that are needed most.
This may seem obvious in hindsight, but I have spent a lot of my career banging my head against the wall thinking "why cant they see this way is clearer" when really what I needed to do was ask better questions, give them concepts to think on and challenge my partners across all roles to consider and share their answers broadly within the organization so that disparate views could be found and decisions made about the future state. They need flexible tools to work with, not rigid visions that seem impossible to achieve.
I find myself putting my hands in my pockets and asking other people to make the diagrams these days. Getting people to consider language and structure, not selling them a better one. That doesn't mean people don't want to BUY it. Which means I have to burst a lot of bubbles about patterns and best practices. But this is the work to me.
I guess I will end with this: there has always been and will always be a need for brave people to ask questions of the language and structures of our world. Without those brave souls, things can get pretty messy.
Thanks Boxes and Arrows (and Christina) for the continued support of this important practice.