In December 2025, John Cleese said: “I want to say I’m extremely sorry that I retweeted some things that I didn’t check properly. I couldn’t believe that some of them had been just completely invented. It was a mistake.” The quote he amplifies below is completely invented.

Few psychology findings are quoted as often, or as confidently, as the Dunning-Kruger effect. And few are so routinely misunderstood. It usually turns up as an insult: stupid people think they’re geniuses, experts are crippled by self-doubt, and the people you disagree with are perched on top of “Mount Stupid.” That famous graph, where confidence shoots up to a peak, crashes into a “valley of despair” and then climbs slowly towards mastery, doesn’t appear in the original paper. Nobody in the study was tracked as they gained expertise. The researchers never claimed poor performers thought they were brilliant, and they never measured intelligence. So a finding about people confidently misjudging what they don’t know is often spread by people confidently explaining a paper they haven’t read. If the effect needed a demonstration, its own reputation provides one. Here’s what the paper found. In 1999, two Cornell psychologists, Justin Kruger and David Dunning, published a paper with a title that has aged remarkably well: “Unskilled and Unaware of It.” They gave students tests of logic, grammar and humour, then asked them to estimate how well they’d done compared with their peers. The people who scored in the bottom quarter had, on average across the four studies, performed at the 12th percentile. They estimated they were at the 62nd. That is overconfidence, but it means believing yourself a bit above average, which is a long way from thinking you’re a genius. They weren’t lying. Kruger and Dunning argued that they faced a “dual burden”: the knowledge you need to get something right is often the same knowledge you need to spot that you’ve got it wrong. If you don’t know the rules of grammar, you can’t see your own grammatical mistakes. The top performers showed a smaller error in the other direction. They underestimated their standing, largely because they assumed that what came easily to them came easily to everyone. They still rated themselves higher than anyone else did, and when they saw other people’s answers, they revised their estimates upward. Perhaps the most hopeful finding tends to get left out. When the researchers taught struggling participants some basic logic, those participants became better at judging their own performance. The blind spot came from a gap in skill, and skills can be learned. The study has limits. It used small groups of American undergraduates, and statisticians have argued since that part of the pattern reflects regression to the mean and our general tendency to rate ourselves as above average. Dunning himself stresses that the effect applies to all of us whenever we step outside what we know. A brilliant surgeon can be badly miscalibrated about tax law. A former banker can make absurd claims about asylum seekers. It describes a blind spot we all have in some subjects, and pinning it on a particular type of person misses the point. It gives us a useful warning: how sure someone sounds tells you little about whether they’re right. The sensible response is to build workplaces, media and politics that reward evidence and track records over confidence, and to keep checking our own certainty while we’re at it. At least I think that’s a sensible response...

First-round others totalled ~9.32M. Both candidates gained ~14M combined in the runoff. That extra ~4.7M came from ~3M higher turnout (abstention fell from 21.08% to 19.2%) plus fewer blank/null votes. Lula took most of the shift. Official TSE numbers check out arithmetically. Musk can't add up.