It beats polls for elections ONLY until someone notices it is being used as the basis of news stories and figures out it will be four orders of magnitude cheaper to manipulate that small market and make the news idiots broadcast that opinions have changed than to actually deploy all the adverts needed to change the opinions.

The very low stakes you point out make this even easier to put a thumb on the scales.

Goodhart's law: "When a measure becomes a target, it ceases to be a good measure"

The point of the article is that as soon as the "news" started reporting on prediction markets or corporatized gambling as if it was a measure of sentiment, it ceased to become a good measurement. That point has long passed.

News already broadcast tons of nonsense. Political commentary is just brain rot. Economic commentary might be just as well generated by one of those Markov chain string generators - it would make as much sense.

>>out it will be four orders of magnitude cheaper to manipulate that small market and make the news idiots broadcast that opinions have changed than to actually deploy all the adverts needed to change the opinions.

It will then become more expensive. Out of all the manipulation news and journalists do every day I am not sure why "people bet money at 1 to 4 odds that Trump takes Greenland before 2027" is particularly problematic. It's true people bet money on it at those odds. How is that more problematic that news running pro or anti Trump segments or broadcasting some random crystal ball readings to justify stock price fluctuations of the day? You can see how much money is bet on the market as well. It's not like you can spend 5 figures and suddenly shape the narrative.