LLM + end-to-end encryption has a long way to go. Hallucinations are other concerns. We’re yet to see useful on-device LLMs. They’ll catch up just not yet.
I spent 45 minutes reading Microsoft's "AI-powered success" blog post (Microsoft advises Meta), and what you described doesn't align with what's happening in the real world. In fact, the opposite seems to be true: more and more companies are using GenAI to help people:
GenAI is a language processing tool, and WhatsApp is full of language-based interactions. But the way Meta has implemented AI makes you wonder: are their managers not creative enough, or is something else going on?
