[natas]

Give it a year or two, and apple or any other hardware will have unified memory OR AMD will have a good offering to run all that stuff locally. It won't be as good as Claude, but it'll do for 90% of the things. It will be expensive as first, just like the first mainframes, then give it another 5 years or so, and it'll be affordable.

[citizenpaul]

I'd argue this has already happened. The highest end mac pro is $10k and down from there. We might be returning to the appliance business model.

Its just a matter of Productizing the software to plug and play light office admin work.

[daemin]

People that do light office work tend to have light office machines, which are very unlikely to have powerful NPUs or even a lot of RAM. Therefore with this minimal setup is it even feasible to do any sort of LLM based work locally on those machines, or will they all be dumb terminals connecting to hosted LLMs of the big companies?

This is also what I wonder, what practical applications can you actually do locally on something like a minimum spec NPU?