Full threadcomplexmango·How soon will we be able to run a GPT4 level LLM on a MacBook, with acceptable latency? 2025?View on HN