I find the realtime translation feature quite compelling. Audio translation requires everything being said twice so doubles latency. If two people from two language speakings both used this, it's a massive step towards a universal translator.

A lot of the other features are meh, reminscent of what a 1980s high tech watch might do for you. feels like a PDA

The part where I get excited are the AI applications, this is also where it gets unethical. i.e. microexpression/voice analysis, AI suggested responses, realtime fact checking of what a person says, and so on. Applications in anything requiring human interaction, sales, meetings, dating, friendships, etc. Imagine being the voluntary puppet of an AI that tells you what to say. If it becomes mainstream, the implicit bias people make towards someone who wears the glasses and therefore may be AI-fed. The inequalities of the have versus have nots, with regards to having such a personal AI.

Ultimately their best bet if shooting for success would probably be to connect it to an AI though. That and social media feeds.