Prediction markets and a reputation system in place of editorial review

Applications
21

Appreciate all the critical feedback. That’s why I posted here first!

  1. The idea is to use a specific crowd - but with enough randomisation within the crowd to create uncertainty over expected consensus - for each piece of news. I believe there would be a Schelling Point within that crowd. It wasn’t hard for a lot of people in Beijing to know if Tianeman Square massacre happened or not. The problem was that the government stifled their voice. It can’t do that on a decentralised platform.

Editors fact check things blind all the time now. They mainly trust their reporters.

  1. I think if reputation alone was so clearly sufficiently efficient you wouldn’t see such different reporting of the same facts in the news today. It’s the owner of a newspaper that decides the context. You also wouldn’t see such sensationalism. Also statistically over 50% of stuff reported is found to be false so that doesn’t strike me as a system that works very well.

People have decamped into tribes and trust the newspaper of their tribe. That’s bad. If you don’t know what tribe members are editing a piece that creates enough uncertainty over the consensus in most situations where the rational bet is to go with the truth IMO.

I agree re subsidies and I have a few systems for them I just haven’t mentioned it here as was solely looking for feedback on the prediction market design. I think prediction markets haven’t existed until now because before blockchains came along they were infeasible from a regulatory perspective. Also early attempts didn’t solve for any of the issues we’ve discussed here. I think my design does so pretty well, especially when compared to traditional news media systems.

We have numerous tools for establishing trust. Reputation is one of them. The law is another. Markets are another. We now have blockchains too. I don’t see why a well-designed system that harnesseses all these tools wouldn’t bring better results than systems that use fewer of them as long as it’s cost effective to do so. Right now it’s just a fact that most newspapers aren’t economically viable. They’re being propped up and there are huge costs to society to them being propped up because they’re gradually losing more and more of their independence. There’s a reason governments across the world are holding hearings on the problem. I believe in taking information intermediaries out of the equation you save on so many costs that you can afford to subsidise a market for truth and the system being much less capturable than it is now. That wouldn’t be hard!

I think if you don’t employ markets I think you just end up with social media which is its own dumpster fire of misinformation. There’s a middle ground here between unviable captured news orgs and the free-for-all that is social media if we harness all the tools available to us to deliver it.

1 Like
22

Murraci, love the idea of a reputation-based editorial system and it’s definitely something that’s possible. A composable set of attestations by trusted parties would solve the problem and make this all possible. Ethereum Attestation Service (a public good for making attestations) provides the base layer for this, as well as the ability to attach payments to attestations via a resolver contract.

Here’s how this could work in practice:

  1. A reporter writes an article on a climate event and publishes it, making an attestation of the article’s authenticity and recording this transparently for others. The reporter could also attest to the sources of information and evidence used in the article.
  2. Multiple specialists are selected randomly to verify the accuracy of the post and make referenced attestations to the article, thereby staking their reputation on the article. The specialists could earn something for this, but it doesn’t have to be financial. It could be social capital or some other form of recognition. This also could have its own attestation that the selection process was followed.
  3. This process starts to build a composable knowledge graph of specialists and the knowledge being shared and attested to, as well as a reputation system for journalists/researchers.
  4. The prediction market could also be created using attestations by resolving real-world outcomes that people attest to, and tying it to a reputation system.

Let’s find time to chat! I’m pretty sure our public good already has a big piece of the tech you’re looking for. Open-source, permissionless, tokenless, free-to-use.

23

The problem with using pure reputation is that people have to decide who the specialists are in the first place. If there’s no cost to that decision collusion and gaming of the system will be rife. It’ll just devolve into what we already have.

There has to be a market element too IMO. Reputation must come from performance in a market. The hard part is getting the market design right.