Just an extended note on that recent @ConversationUS / @sciam piece claiming to debunk the Dunning-Kruger effect. It holds the usual pretense that research on the DKE stopped with our 1999 paper, implicitly suggesting our Study 2 is the only relevant evidence. 1/n Doing so, the piece cherry picks data it wishes to discuss, focusing on one study out of dozens that have been published ever since. This is the typical move of critics from this genre, 2/n

Thread: Spending time this week exploring overconfidence in AI systems doing machine learning. It’s a thing: AI systematically overestimates the chance its judgments are correct. Fascinating. Leading causes: 1/5

forbes.com/sites/kareemsa… 1.Machines have trouble when new stimuli range outside of what was in its training set. A system set up to tell cats from dogs categorized the researcher’s human face as a dog with 98% confidence! 2/5
jramkiss.github.io/2020/07/29/ove…

Are the “rich, powerful, or famous … more likely to fall prey to Dunning-Kruger.” I’d say “no.” We all are vulnerable and just as opinionated, but some of us are more likely to have our opinions broadcast over cable TV, newspapers, and social media. That said… 1/n

https://twitter.com/work_matters/status/1252027068562370560

people with some training that has illusory relevance to public health may misapply that training unwisely into epidemiology. A mathematician will plot growth curves, missing important known facts about disease spread. … 2/n

Reminded today is the 20th anniversary of the NYTimes article that started all the Dunning-Kruger hullaballoo. Apologies for the indulgence of this thread. #DunningKruger 1/n
nytimes.com/2000/01/18/hea… Erica Goode @egoode of the NYT had called me up a month earlier, having read the title of Justin’s and my 1999 paper. Could she call me back after she read it? “Um, sure…” 2/n.
avaresearch.com/files/Unskille…