Bonus: This is a custom chatbot based on ChatGPT. It brands itself as a „companion for your mind“ (heymia.ai). What is the stance? What is the conversational project here? Turns end in multiple questions like throwing spaghetti at the wall to see what sticks. What does it do for the conversation?

This is Claude 3.5 Haiku. Anthropic clearly still works in "generator" mode. What does its rational stance do for the conversation? How does it orient to the (simulated) actions of the human?

ChatGPT - whole sequence. Turns are a lot less "chatty", since OpenAI also provide a "voice mode" for their chatbot (which sounds creepily human by the way).

A virtual shoulder to lean on?
What does it mean for a chatbot to give comfort and support? How do ChatGPT and Claude handle opening with a problematic pre-telling? See the whole sequence in the thread. Thoughts? #emca #openingsAndClosings #GenAI #conversationality

Enjoying myself immensely watching the #LboroCADay24 having a very British moment complete with tea cosy and cookies.

Dear, Apple Intelligence! Interaction does not deliver information, it delivers action.
(slightly rephrased Levinson 2010).

Merely summing up the "information" is not helpful. AI needs to understand social behavior. #emca FTW!
From CNET, seen on YouTube: www.youtube.com/watch?v=D0V5...