@RLanceMartin posted this under the main article, but I read this and thought it was really insightful! I'd love to see your exploration of Search vs Memory vs RAG and combinations of those (like just Memory + RAG vs Search + RAG vs all three together). Maybe even RAG applied to memory itself?
@trq212 One sketchy part of this is that you’re potentially copying over and saving (at least for the duration of the session) a lot of my very personal info and files directly to your VMs
@trq212 Yeah I can’t bother doing certain things anymore now that AI can do them for me. On one hand I can parallelize better, on the other I’m way lazier when it comes to certain tasks
@paulg Models can definitely become and already in some ways are dangerous. But how did the 3 dudes who’re known to dislike each other make peace overnight? Whatever the answer, it’s not good
@KianSadeghi5@nucleusgenomics How’d one know this only increases intelligence and doesn’t sacrifice something else? Human qualities and abilities do not develop in silos and it’s the combination that yields truly outstanding people
@ThePrimeagen Your arc from being skeptical of agents to making them do human tasks like play video games or browse internet and making agent tooling is super cool to see