AI coding dream team of agents for VS Code. Claude Code + openai Codex collaborate in brainstorm mode, debate solutions, and synthesize the best approach for your code. - DeepMyst/Mysti
If on machine is unreliable, throw more unreliable machines to the task, and all be resolved. This is probably the most frequently advertised strategy.
Only tangentally related, but lots of things in tech work this way, from RAID to load balancing. But they’re doing well defined tasks with programmable control, not fuzzy llm.
If on machine is unreliable, throw more unreliable machines to the task, and all be resolved. This is probably the most frequently advertised strategy.
Only tangentally related, but lots of things in tech work this way, from RAID to load balancing. But they’re doing well defined tasks with programmable control, not fuzzy llm.
I didn’t work with highloads, but ye, if I think about it, I have to admit there’s a time when that strategy might be useful.