1. I am open to the idea of two days but, think it would benefit to being less hours each day than the original summit. 2. I'm happy to share what I have learned designing for AI driven experiences. (If I learned enough in my contract role)

LLMs can def miss very important caveats or upsells in documentation or other places. That means we need to find another way to communicate with users outside of the LLM or we need to be able to provide content to LLMs as "author's notes" when LLMs ingest content. This might be helpful in more ways that one.

Update: I was just offered a contract role, for a company that i've worked with before to bring AI to both the design process and the product itself!

It's an interesting concept. I'm wondering if any verification would be needed, as an AI generated interface can create misunderstandings depending on the visual hierarchy. Like even if the UI morphs based on the context, is there some way the decisions made can be shown in standardized way so the user knows what they are doing? At least in enterprise products very specific decisions on the UI are made to showcase the many nuances of the algorithm/product/decisions. I created this tool in Figma make in about 20 minutes for a pain point i've had with Delta airlines it allows me to easily downgrade myself without calling an agent. https://evoke-meteor-41742295.figma.site . now i'm wondering would that be something this generated on the fly or highly customized or curated? that's an interesting decision, only plat/diamonds can use this tool as no one has upgrade certs. but why would the UI need to be generated more than once, and i'm sure legal would want to review all the copy and product marketing would want to ensure there isn't anything confusing on there that might end up backfiring and causing more support issues. Sure it's the direction we are going, but maybe it's a bit overblown how comfortable companies and cultures will be to release agency to the AI for very narrow use cases that can have a big impact.