Comment by Zecc

7 months ago

> Once (a) models are capable enough to distinguish between semi-plausible garbage and possibly relevant text