But if we take it back to how people are starting to interact with these tools, and let's stay on large language models while we're here, what's the UX conversation around that right now happening in agencies?

Yeah, well, I think there's two sides, right? There's LLMs as a tool.

So how do we use it within an agency to make our workflow faster or create a better experience for the client and the end customer? And then there's also as a consumer product, What's the experience like there? What type of experience are you trying to create? So on that one, for example, we're working with Chris Reese Edwards out of the Queensland University.

He's working with a digital well-being lab there.

So he's really looking at what is that.

We don't want to character AI type, do anything for you and be an AI girlfriend or whatever.

and we don't want claude but what's that middle ground where an llm could potentially be empathic but also be able to detect when uh you're not in in the right position right and can refer you to appropriate services um and chris famously worked on a project or he was the ceo of a company called soldierly which looked at detecting ptsd in army veterans and referring them to appropriate services.

Talk us through what is spatial, how is it defined right now, and how are agencies going to struggle to keep up with the future?