Oddbean new post about | logout
▲ ▼
AVERAGE_GARY | 1 months ago (raw) | export | reply | flag +6 
 Waiting for local LLM is new coffee time for devs. Used to be compiling.
▲ ▼
arkinox | 1 months ago (raw) | root | parent | reply | flag 
 I was just looking at setting up ollama today. Is this de wey?
▲ ▼
AVERAGE_GARY | 1 months ago (raw) | root | parent | reply | flag 
 Meh. Depends on machine. Im running a M3 MBP and it's fast enough. ChatGPT is probably better but I don't use it
▲ ▼
Nathan Day | 1 months ago (raw) | root | parent | reply | flag 
 I tried local (Ollama/Llama) with Code plugins, but I'm finding cloud service like replit much better.

Whole codebase context and much faster.

Replit agent is nuts.
▲ ▼
Nathan Day | 1 months ago (raw) | root | parent | reply | flag 
 Was thinking exactly the same today whilst waiting on my AI slave.