“waiting for deep research” is the new “it’s rendering”

Farcaster-native research: https://paragraph.xyz/@matthewb | prev. Juicebox, Nouns, RAW DAO | casting about coffee, hi-fi, and other rabbit holes

neuroscience phd, DeSci advocate, Buddhist chaplain. co-host of the /okbanger show and vibe imagineer for /christin. https://christinchong.com

in my experience, the occurrence of those failures has gone down over the last 1-2 months.

not sure there’s much you can do to mitigate it, I just babysit it until it actually starts researching. occasionally it gets stuck at the start if I switch focus too fast.

do you find this happening often with o3 or which model?

how do u prevent ChatGPT deep research from crapping out? mine does it sometimes and it’s unpredictable what causes it (likely “stuck” on some LLM trap from cloudflare or something )