Decentralized Media Via Web3: A Solution to Bias and Trust Issues in News? – CoinDesk

The Times was “unabashedly pro-Hitler in The Thirties, serving as a sturdy fount of Dr. Goebbels propaganda” and was “unabashedly pro-Stalin in its coverage of the famine in Ukraine,” writes Rindsberg in “The Gray Lady Winked,” who then elaborates on a series of charges. “You’ve got a single family in control of this newspaper for 120 years,” Rindsberg said to me in a recent interview. “It’s a dynasty. Their interests and their incentives are so misaligned from their reader’s interests.”

I’ll admit I’m a skeptic; like most journalists I enjoy and regularly read the New York Times. I’ve written for the paper. I’m hardly unbiased. (And to be more explicit, these allegations are not ones I personally endorse.) But whatever you think of Rindsberg’s specific allegations of the paper, there’s common ground in the idea that, generally speaking, there are fundamental problems in publishing that no one has figured out how to crack.

Rindsberg proposes a solution: Decentralized Media, or “DeMe.” What if citizen journalists were empowered and incentivized to fact-check, do more analysis and make sense of raw data and the world? Rindsberg suspects Web3 could fuel this. “The technological solution is definitely primed for blockchain,” says Rindsberg. “To create a chain of custody for facts, claims and sourcing.”

We might not agree on everything, but the two of us had a lively and respectful conversation about his critique of the New York Times, the structural problems with the space, his vision for decentralized media,and why he thinks that traditional publishing is “optimizing not for objectivity and not for ideals, but for results, page views, clicks and subscriptions.”

I think it’s both, in a way. There are some mechanical errors that you see, like when sourcing doesn’t actually have a source, and you see that everywhere. But the New York Times is just bigger, so in a sense it all gets magnified.

But then there’s the ownership structure. You’ve got a single family in control of this newspaper for 120 years. It’s a dynasty. Their interests and their incentives are so misaligned from their reader’s interests and what their own reporters’ interests and incentives might be, that the divergence is really stark.

Sometimes people ask me, “If the New York Times did all this crazy stuff, how come they were able to retain this number-one spot in news?” My response is that it was because they did these things that they stayed number one. It was because they had a pro-Nazi journalist in Berlin in 1939 that they got unbelievable access to Nazi sources that got them the best scoops. There are many examples like this.

So, who’s done it better than the Times? I don’t know that there’s anyone, in a general sense, who’s done better. I know that for the book I would frequently go back to look at the Washington Post reporting on the same issue at the same time, just to compare. And the Post was generally pretty sane and commonsensical on most of these topics.

Oh, I read the New York Times every day. I’m still a subscriber. And if they were ever to ask me to write for them – I don’t think it’s gonna happen – but I would say yes.

On a more mechanical level, that means using bad tactics like bad sourcing or selective sourcing or the bias of omission or the bias of commission. They’re increasingly doing this to advance an agenda – and that could be financial, it could be political or they are doing it to advance a thesis. With journalists, the temptation is always there, right? You come in with a hunch, you come in with a great idea. But as the standard of objectivity is loosened in the news media, these types of practices are encroaching on neutrality, and disinterestedness, and dispassionateness.

And part of the problem is media concentration. We have basically 90% of the news media in the hands of six companies. That adds fuel to the fire because we are optimizing not for objectivity and not for ideals but for results, page views, clicks, and subscriptions. These conglomerates obviously have an overweening financial objective. They have a fiduciary responsibility to their shareholders. And that means the news organization has to deliver, come what may.

Was it necessarily this exchange of goods? I wouldn’t go that far, but it’s gonna influence you. It’s gonna influence reporting. Either way, we get the myth of SBF. It took two years to build it. And it took two days to burst it.

And that was done through pretty much two, maybe three outlets: CoinDesk, Richard Chen on Twitter and possibly a Substack account or two. That’s decentralized media. I mean, CoinDesk is closer to centralized media, but it’s carrying the themes of decentralization.

Maybe it’s not a tipping point but it’s a real indicator of change. What does decentralized media look like in the future? I think it’s a lot about changing how we understand reporting. From narrative editorial to something much closer to data interpretation and visualization, but interpretation by the audience. To say, “Here’s the raw data, and here’s the structured data, you audience members gather together on a topic of interest and generate the visualizations. You draw your own interpretations and conclusions from the raw data and put it out into the world in social media.”

And then there’s what I’m doing with this company that I’m building, called Alitheum, which is giving people the power to understand the reliability of any piece of news media. To take an article or a paragraph and then understand how much bias it has. That’s also the piece of this puzzle – decentralizing media by giving people the power to decide for themselves.

So then you don’t have to say, “The New York Times is not reliable.” I never have to come to that conclusion. I can say, “Oh, this article meets my reliability standards but that one doesn’t, and I don’t need to throw the baby out with the bathwater.” This boosts my trust level with the New York Times because I verify. That’s something we constantly see in Web3, that trust is one thing but verification is another. Don’t just trust, verify.

The technological solution is definitely primed for blockchain. To create a chain of custody for facts, claims and sourcing. That’s one of the biggest problems; a New York Times article could link back to its source, but you don’t know where the next layer goes and where did that source get it from? And if you go back far enough, did it come from some incredibly dodgy or racist?

So to be able to create that chain of custody, I think that’s incredibly important. To be able to validate that people are who they say they are. Or that a journalist has validation that [information] really did come from a certain person. But I think there’s also a verification of reliability.

At Alitheum, we’re putting articles into a system where it can say something like, “It used named sources eight times and unnamed sources two times, and it only used a false source once. So we get a rating of 76%, and here’s why.” Or we can look at tonality and how inflammatory or biased the language is.

I’m working with a professor in Zurich who is developing very reliable computational models to measure bias in news. You feed an article into the [artificial intelligence] and it will tell you how much bias is in the article.

So, what Citizen does for monitoring crime, right? If there’s a crime in your neighborhood, you get a ping. That’s already gone half the distance. Then go a little bit further and say, “OK, bring that data on-chain. Now it is actually verified.” You can bring it from multiple sources and multiple nodes the way that Chainlink does. Now we have an event that has been turned into data in a way that is verifiable and also that is decentralized, so it’s not coming from a single source.

Yeah, I think there’s some kind of an in-between. You have the Substack model, where you’ve got people who independently say, “I want to cover stories of the Southeast Denver metro beat, which has disappeared completely from the local news because they can’t afford it.” But if every time you want to report on some small or mid-level crime, it’s going to take you the whole day to do the story.

But if you’re able to go source the data from your desk and then turn it into a visualization, then you become something like a Substacker/influencer. What we need is a way of gathering information for you, as a single individual, to create meaningful content to feed the audience with relevant stories. This is not about you going to interview four policemen and the victim; that’s not possible with any single individual. You can’t do that. So this is a way of solving the supply side [of data] of this news media ecosystem, where we have independent individual creators or journalists.

And there’s also a model where those types of people start to clump together. Like the way that Bari Weiss has done with her Substack. Maybe it’s not just you focusing on Southeast Denver, maybe it’s you and a small group, and it’s not just Denver crime but also Denver sports and Denver business. But it’s all based in data. It’s not about you pounding pavement, it’s about you pounding data.

And there are models that could even support fact checking. If you’re able to verify a fact in a story, then a way to reward that would be through rewarding tokens. I think that’s a fantastic way to do it.

But I think when you look at something that is very prone to error, even if it’s on the “human” side of the spectrum, like in a battle where you’ve got the fog of war, nobody’s able to make sense of what’s happening. A journalist who’s cowering behind a rock trying to figure it out has zero chance. How do you piece together what actually happened without making a significant error?

What if with the internet-of-things, with sensors, with geolocation you’re able to track troop movements and then pull that data together and assemble it into a picture of what actually happened in that battle? You don’t have the fog of war, you have the clarity of data. I think the internet of things will be such an important part of that. This may enrich some of the more traditional, narrative-based reporting as well.

In five years, we’re going to be seeing a lot more visualized data, a lot more charts and graphs. We had this spike of interest in data journalism with Nate Silver [founder of Five-Thirty-Eight], and it kind of got absorbed into the larger field. I think that’s going to rise again in a much more significant way.

And not just for politics. We’re seeing that in sports now, where it’s becoming more data-driven. The [general managers] of sports teams are now data scientists. But on the media side and on the reporting side, it’s still “this player sucks or this player is great.” People will want to understand the data as a first resort, not the last resort of analysis.

It will be about connecting the offline world to the online world. Such as using sensors. And I think that’s probably going to involve Web3 and blockchain in some significant way, so that in an automated fashion we can show that a certain chain of custody is not broken, or that we understand that events have been verified because of the way that they were brought on-chain. And lots of on-chain data will be analyzed as a kind of news. How that will look, exactly, I’m very curious to see.