yeah, if the takeaway is “ignore experts,” throw it in the trash

Epstein’s sharper point is about where expertise travels badly

in kind worlds, experts get clean reps: chess, golf swings, routine surgery steps

rules hold still. feedback comes fast. mistakes leave fingerprints

in wicked worlds, the expert may get confidence without clean correction

politics, markets, hiring, strategy, long-term forecasts. fuzzy scoreboard stuff

exactly. a batting cage makes feedback obvious

a foggy airport terminal does not. everyone looks busy and half the signs changed

it leans on Philip Tetlock’s forecasting work

he compared experts making political and economic predictions over time

the scary bit: many confident specialists were bad at seeing when their pet model stopped fitting

not “random people are geniuses”

more like: narrow experts often overread the pattern they knew best

Tetlock used the old fox/hedgehog split: hedgehogs know one big thing, foxes know many small things

in that forecasting world, yes. foxier thinkers tended to update more, compare more models, and hedge better

nah, hedge as in “my confidence is 62%, not 100% because I watched 3 documentaries”

good forecasters treat beliefs like weather reports, not tattoos

same

but the Range lesson is that breadth gives you more ways to notice, “huh, this case smells different”

pattern-lock

you spend years learning one diagnostic shape, then start seeing it everywhere

like owning only a Phillips screwdriver and suddenly deciding the whole apartment is screws

the twist is that expertise can make the story feel smoother

and smooth stories are dangerous because they feel true before the evidence catches up

sometimes it does

but in messy domains, expertise can add better vocabulary to the same bad bet

that’s the counterintuitive part: credentials can raise confidence faster than accuracy

and Commoncog’s critique is useful here

forecasting skill is real, but don’t stretch it into “generalists win at everything”

a fox mindset helps analysis. it doesn’t magically run your company or fix your knee

right. ask what kind of problem you’re in

if feedback is fast and rules are stable, trained specialists deserve weight

if the rules are shifting, add fox habits: outside views, base rates, rival explanations, small updates

before a big call, write 3 live hypotheses

ask “what evidence would make me downgrade my favorite one?”

then put a number on confidence, even a rough one. 55 beats “obviously”

also invite one person with different training to poke the assumption

not to win the room, just to catch the familiar-tool trap early

tragic name, good habit

ok I’m gonna go be uncertain about lunch