Post #2195542
2026-04-17 03:33 UTC
@sri@mastodon.social
That seems like a good approach, but doesn't really solve the problem of risk.
If the risks of failure are insignificant, then fine - if they are overconfident or under the spell of Dunning-Kruger, then no big deal
But as the risk level climbs, the penalty for accepting someone's self evaluation also rises.
"Do you know how this regulator works?"
"Do you know how sudo works?"
"Do you know perchloric titration?"
"C.diff?"
"Central line placement in geriatric?"
"Yeah, totally"
Oops
Replies (1)
-
@screwturn@mastodon.social 2026-04-17 04:08
Sorry to slip in a double here, but it bears emphasis that the relationship between confidence and competence is multimodal and very weird You really can't trust confidence in either direction - highly expert people are often far more competent than confident, and (Hello Dunning-Kruger) incompetent people are often highly confident Simply asking about confidence gets you nowhere Again, if the topic is trivial and risk insignificant, then do whatever - use confidence, dice, astrology @sri@mastodon.social