Prediction Markets for News
dyokomizo opened this issue · 0 comments
News is fake news nowadays. Corrections ain't.
At best journalists care about not intentionally misleading people, but we want to know what is true and not what their sources told them.
Prediction markets can be used to assess how reliable are media outlets, journalists, or even specific subjects.
News articles can be analysed and broken down, maybe using something like Argdown.
Similar articles can be grouped together and relevant differences can be singled out for discussion.
Individual claims/arguments can also be bet on through prediction markets. We can derive the market confidence for the whole argument.
Claims/arguments can be connected across news articles, so later evidence and bets can back propagate and allow people to review past arguments and update their beliefs more explicitly.
For me objectivity is tied to epistemology, it's about having beliefs backed by evidence, updating your confidence on them given new evidence, and being able to reason counterfactually.
Belief becomes evidence, by transitivity.
News don't disclose how confident they are about the evidence they show.
As we try to assess how much confidence we should assign to news, we end up with low values because they showed us so many falsehoods.
Also, much of news is implied, which is its own can of worms.