The Stack Archive

How the blockchain could save journalism – for free

Wed 7 Sep 2016

Over at hack.ether.camp, an annual four-week online event which is a hybrid between a hackathon and a startup incubator, participant developers suggest abstract ideas around which others may gather and build momentum. Today, one of the contributed ideas from mid-August appeared in Hacker News, to little more acclaim – so far – than the idea seems to have received at the Ether Camp forums.

And I wonder why, because, frankly, it is brilliant. The only problem is that in doing the world a favour, it would probably wipe a few hundred billion dollars off the online publishing market, since it would completely undermine the curated and automatic recommendations systems which are the central financial pillars of Facebook, Google, and a handful of other influential global internet players.

Its ramifications go far beyond its initial suggestion, but for the record, here it is:

Do you trust what you read online?

Real journalistic investigations are very important, as they uncover great injustices, corruption, and pollution. But they are suffering under the economic pressures of money, politic and click-baits.

I believe that we can solve this problem by building a global decentralized unstoppable news source with reputation system for investigative journalist and payment directly to the reporter

I said ‘free’ in the title, didn’t I? And yet this post refers additionally to micropayments, which is, pardoning the pun, old news. Not to say completely irrelevant to the central idea presented here – that a blockchain-based reputation system, divorced from the potential prejudice of human-curated systems and the fallibility of the automated systems that are trying to replace them, would automatically bring power – and, inevitably, increased remuneration – to the person in receipt of high global ‘trust’, and to the enveloping organisation around their work.

If this simple idea has a flaw, it’s that it has been overthought. A blockchain-based reputation system with wide participation is not only adequately ambitious, but enough to achieve the poster’s stated objectives. It’s enough to change everything.

The commercial version

The idea of a ‘global’ and ‘decentralised’ news force has appealed to at least the two biggest players in online publishing and search/social-driven traffic. Facebook’s Instant Articles feature, instituted in 2015, has engaged a slew of major global publishers which have been struggling with maintaining the viability of their own hubs in the face of the rising social market share, and with the increase in adblocking among end-users.

At the same time Google has been experimenting with direct publishing to search – a tacit threat to Twitter – and with its Accelerated Mobile Pages Project, which also seeks to speed up delivery by cutting out the middleman.

Well, all the other middlemen.

And this is the flaw with any ‘vote-up’ system which has to run servers, maintain extensive development or employ staff; curation systems such as Google, Facebook search, Hacker News, Fark, Pinterest, Slashdot, LinkedIn, Reddit, Twitter; user focus simply shifts from one locus of self-interest to another, one set of ethical or commercial biases to another.

Voting up the reputation of an individual investigative journalist – or an author of acknowledged merit – via a blockchain-based system of ‘reputation credit’ would distribute the logistical challenges across all the people who would benefit from the system; which is another reason to keep it simple, and to stick to unrepeatable blockchain vote-ups, both for individual contributors and for individual articles or works. Where they reside is irrelevant.

Object-oriented praise

On a practical level, individual contributors and articles would need to be defined as uniquely hashed entities, a native function of blockchain systems which can (usually) only go wrong in one place – the start. So a blockchain reputation system would need to develop a form of consensus identification for an entity, wherein an individual becomes the central node in a fractal branch of his or her own work.

With the central entity defined – in this hypothetical example, the author – new works attach themselves to the writer as they appear, with article upvoting raising the writer’s reputation, and reputation votes raising the writer’s output.

Delivering the vote and maintaining the ideal

In the age of the swipe, no-one is going to run a command-line install for a scheme with communal (i.e. marginal) resources. So if a blockchain reputation vote were to be as simple as a browser-based ‘Vote up’ button, it would either have to be built into a popular browser (and let’s face it, only Firefox has the neutrality to consider it) or implemented as an add-on or plugin – and run the gauntlet of becoming a popular install, and being a (presumably) unwelcome guest in the Chrome store.

No matter how autonomous a blockchain reputation system is rendered – perhaps with any web-facing aspect also decentralised – at least a few real people would have to commit themselves to looking after it, and to providing monitoring and security services.

And the more people become involved, the more distilled the scheme’s objectives are likely to become. If it remains marginal or difficult to deploy with ease, the guardians are likely to disappear; if it were to achieve its true potential, it seems they are likely to become corrupt and avaricious.

So perhaps this isn’t the year, or even the decade, in which an idea as socialist as the Ether Camp suggestion can come to pass. But anyone who dismisses the notion has little right to complain about the proliferation of clickbait or the escalation of the adblock wars. If all we are going to do with the blockchain is watch corporate entities strip-mine its featureset, in spite of its enormous potential for global emancipation, we’ll just have to keep wandering from one proprietary and self-interested point-and-click hub to another.


adblocking advertising Blockchain cryptocurrency feature research
Send us a correction about this article Send us a news tip