@DorianDevelops Opus 5.5 on high
3-4 Active sessions
1 doing deep architectural research
1 auto building simple well defined features
1 prototyping new ideas
1 iterating on model improvements or prompt engineering
a weird inversion with LLMs is the models improve faster than the tinkerers
when i see people with custom workflows and setups they're all addressing problems that don't exist anymore
the person naively using vanilla codex is more likely to be experiencing state of the art
@ImBlake How do you see this manifesting? Do I get an allotted token quota to use on my device? Or is the AI provided through devices embedded in the unit?
@ImBlake I haven't, but the thing is it's not just about the model, its the UI and harness. Session management, code diffs, tool integration, small UX sugar. It has become an AI first IDE.
Codex is pretty great too but I generally have a better experience in Claude Desktop.