Zum Seitenende Übersicht Artikel Home & Impressum
First the link to this week's complete list as HTML and as PDF.
***
Lazer et al. take note of a vexing problem: Making media consumers more aware of fake news will result in them becoming more critical of and less prepared to accept state sponsored propaganda and the unified promotion of political correctness. It seems the old monopolies just can't win this one.
Speaking about Facebook and similar platforms they note that those could place more emphasis on quality and reliability in their ranking. But why should they? As Lazer et al. state correctly “They use complex statistical models to predict and maximize engagement with content.”
That's what they sell and what they earn their revenue from. Facebook users are not the customer but the merchandise. If you wanted them to serve you, you would have to become an attractive customer and be prepared to pay for their service. Selling their soul for some supposedly free entertainment is everybody's own choice, but complaining afterwards is just plainly immature and stupid. Everything Vosoughi et al. have found out could trivially have been predicted by spending two minutes thinking about what Facebook is, does, and where their revenue comes from. Next week I expect an article in PNAS boldly claiming to have proven that two plus two make four.
***
There still is some hope left. Both Avino et al. and Cesana-Arlotti et al. give us real data points behind their regressions. In the case of Avino et al. the use of the standard error too is correct. This study is not medical, trying to make a prediction for the individual case, but about differences between groups.
Zum Anfang Übersicht Artikel Home & Impressum