Haven't seen Mozilla's LocalScore [1] mentioned in the comments yet. It's exactly made for the purpose of finding out how well different models run on different hardware.
Like the other comment already said, the search is still Retrieval, and putting it into the context window is Augmenting, and the LLM request leads to the Generation. A.k.a. RAG.
Yeah, it's funny to see them going in opposite directions. Then again, Squarespace is operating at a far bigger scale than Motion is, so it kind of makes sense. I partially believe them when they say that "we had already pushed vertical scaling to its limits". My guess is that continuing to use Postgres would have been totally possible, but perhaps much more annoying.
> I am not a mac person, but I am debating buying one for the unified ram
Soon some AMD Ryzen AI Max PCs will be available, with unified memory as well. For example the Framework Desktop with up to 128 GB, shared with the iGPU:
https://x.com/michpokrass/status/1869102222598152627
It says:
> hey aidan, not a miscommunication, they are different products! o1 pro is a different implementation and not just o1 with high reasoning.
reply