Yeah, though most people weren't even running local LLMs, they just heard you need a Mac Mini to run OpenClaw so they got one.

I think it might have been a mix of the two to be fair, but probably a lower portion of devs. For example, I have a Macbook Pro with an M4 Pro but in about two more years once I've put aside money for a newer laptop I'll wait for the next "updated" models and buy another Mac just for the inference, the speed improvements and the unified RAM.