Unless this speaker runs a local model somehow, what makes it different from a speaker with the ChatGPT API integrated? Is that all this is?

Edit:

The device — described by suppliers as something that resembles a screenless smart speaker (in other words, an Alexa-adjacent device)

That sure sounds like that's all this is. With how many OEM parts related to smart speakers there are available these days, doesn't 20+ engineers seem like a lot for something like that?