Anyone will run LLMs locally soon on about every piece of hardware. If it’s part of the system like the network stack / firmware, now does your mental framework

Anyone will run LLMs locally soon on about every piece of hardware. If it’s part of the system like the network stack / firmware, now does your mental framework work. Progress is murdering their business model. They are trying to stop it; which makes sense, but let’s not pretend the trajectory here isn’t to ubiquitous LLMs everywhere within 3 years. We have to plan for that, assess the benefit for society from that.