Claude Opus 4.8

Re:Claude Opus 4.8 by binary0010 on Thursday May 28, 2026 @01:16PM (#48312205)

Maybe try making a simple randomize script to swap the three latest models. And see if you can tell which ones are meaningfully different without knowing which ones are flipped on or off?

[ Parent | on HN ]

Re:Claude Opus 4.8 by osigurdson on Thursday May 28, 2026 @01:34PM (#48312493)

I find the quality ebbs and flows even on the same model. My guess it is something to do with GPU availability but only guessing.

[ Parent | on HN ]

Re:Claude Opus 4.8 by atq2119 on Thursday May 28, 2026 @01:58PM (#48312868)

Unless you're systematically repeating the exact same task, the most parsimonious explanation is that you're seeing natural variation based on different tasks, random sampling of tokens, etc.

[ Parent | on HN ]

Re:Claude Opus 4.8 by osigurdson on Thursday May 28, 2026 @07:12PM (#48316835)

I don't think this explains the phenomenon as is more temporal in nature - not prompt to prompt. I'm sure the AI labs gracefully degrade to simpler models when resources are low - why wouldn't they?

[ Parent | on HN ]