IS THE ECONOMIST ALWAYS WRONG? Scandalously, in some circles @TheEconomist has a reputation as a contrarian indicator. This week we fessed up to getting a big call on oil prices from April wrong. Obviously our goal is not perfectly-hedged (and perfectly boring) predictive accuracy: often it is to stimulate, provoke, and challenge. But I did want to test that wider allegation, so I ran a series of LLM scorers across our full leader database since 2000 (7,000 leaders in all.) You can see the results in the chart below: each dot is one of the 1,400 leaders where we identified concrete and falsifiable predictions that were central to the argument. Higher = more accurate, further to the right = more contrarian. We do well, unsurprisingly, when aligned with conventional wisdom. We often do worse when truly out on a limb. But actually, on average, we are a bit likelier to be right than wrong on our somewhat-out-of-consensus calls. All round, a respectable performance. And as @ecurrnomics points out an accompanying leader, there is no shame at all in being beaten by the market: as good free-marketers we believe deeply in the aggregated wisdom of prices. Take a look at my piece here, which includes a canter through our best and worst calls of the last quarter-century: economist.com/interactive/fi…

In a 3% experiment, removing the Top-30 highest paid revenue share accounts from the For You timeline increased both time spent and daily active users on X.

Humanity's ability to know, reason, judge, and act well is the foundation of science, democracy, crisis response, & management of AI itself. AI poses serious risks to that foundation. New paper on epistemic risks by 30 experts calls for attention to this. Link in thread.

Feedback Loops: Human-AI and AI-AI feedback loops are narrowing the epistemic space from which humans and AI draw. This already drives homogenization, and may lead to fragmentation and more self-referential information environments.

Political identity drives choice of large language models—even when accuracy is incentivized. Participants (N=1,884) quickly developed preferences for AI systems that aligned with their political identities, and these preferences were stronger when models carried recognizable brand names. In the second stage, 71% persisted with their previously preferred model despite incentives for correctness. This reveals that users do not treat AI systems as neutral tools. Instead, they select between them in ways that reflect political identity. osf.io/preprints/socarxiv/z5… This is consistent with the identity-based model of belief: People select information sources and allocate attention toward in-group sources. You then need strong incentives to override their partisan bias: sciencedirect.com/science/ar…