“Fake News” appears to be an unavoidable plague of the rapid connectivity digital age. Take any issue of wide-spread concern and you will find a variety of opinions:
- Gun Control
- Climate Change
- Economic Policy
- Legalization of Cannabis
The list goes on and on. At least in the United States, the First Amendment to the Constitution provides all citizens with a right to freedom of speech. That right can be curtailed somewhat, but in general, anyone can say anything about anything. Unfortunately, there are those out there who take topical advocacy a bit far. There are those out there who will actually make false factual statements, “Fake News” designed to sway public opinion.
On any matter of any import, once released into the hyper-connected world of social media, a well-designed bit of Fake News may achieve a viral spread, causing massive popular disinformation. Responses to widely distributed Fake News may follow, but in general, people will self-select and believe what they are most exposed to. So, the factual elements upon which the voting nation forms its opinions donates its money and casts its votes are no better than a sandy, unstable foundation. What good is voting if you don’t know the context around the subject of the vote?
A Centralized Lie Should Not Withstand a Distributed Factual Analysis
Bitcoin and other distributed ledger technology protocols have a singular primary function. They provide truth. Specifics may vary, but in a majority of cases, the distributed ledger paradigm is designed to assure truthful behaviour by incentivizing a large number of stakeholders to collectively validate and record facts. For Bitcoin, the “facts” are that the person sending a Bitcoin has possession of the Bitcoin and has not previously sent it elsewhere. If a person tries to send a Bitcoin that they do not own (a lie, equating to “fake news”), the blockchain functions to reject the transaction.
The important thing here is that the validators all have copies of the blockchain (ledger) to examine when validating a transaction. In other words, the validators are already holding all of the “facts.” If we want to use a distributed validation paradigm to defend against fake news, we need a large number of validators that all have access to the real facts. That’s not easy. Transaction history is linear, limited in scope and most importantly, created and stored entirely on-chain. The moment that we introduce off-chain data to interact with a distributed ledger system, we run into a conundrum known as the “Oracle Problem”.
How do you know? Ask some oracles.
An oracle is a provider of data. For our purposes, a provider of facts. The facts that an oracle provides are (for our purpose) not originated as part of on-chain functionality. For the most part, oracle solutions are discussed in the context of financial products. That makes sense in that the direct repercussions of untrustworthy financial data (ie. token prices for a decentralized crypto exchange or results for a betting market) would have serious immediate negative consequences. However, oracles can and should play a much larger role in bringing the benefits of decentralization to multiple industries. Corresponding to our topics above, one or more oracles might collect data like:
- Gun Control: Number and type of incidents of gun violence in particular municipalities
- Climate Change: Weather data provided to universities via satellite transmissions
- Economic Policy: Data provided by IoT sensors
- Legalization of Cannabis: Use of medical services by persons who use particular cannabis products
But how do you know you can trust the oracles.
A blockchain’s data integrity is dependent on the sample size of it’s validating constituents. Without a sufficient number of independent oracles, bogus or malicious data can perform the equivalent of a 51% attack. In other words, in order for oracles to serve their purpose, the process of filtering data through the oracles needs to be as decentralized as the process used for validation of on-chain data.  The data provided to the oracles needs to be independently sourced from primary materials so that the oracles agreeing that a fact is “true” is more than a case of all of the oracles seeing the same Fake News.
The truth doesn’t come easy.
The fake news problem may appear unmanageable. Content aggregators (ie. Facebook, Twitter, Reddit) must now validate the truth of user posts or face the wrath of lawmakers and the public at large. Their centralized approach (each being responsible for the truth of facts that their users disseminate) cannot handle the volume and complexity of the data. Examining content for truth manually would be extremely expensive and largely inaccurate. An automated examination may be more practical, but typically algorithms can be gamed fairly easily.  On top of that, the truth is not binary, shades of grey will introduce subjectivity into the process. One can predict that given the public pressure, the Facebook/Googles of the world may block all but the most verifiable facts. That result would eliminate some of the prime news use cases that the internet has to offer.
Building an oracle based solution is not easy either, but it is theoretically possible. Consensys attempted to build an oracle solution product but encountered some difficulties along the road.  The Chainlink project, establishes nodes that aggregate multiple sources of data to ensure that a single source of data (which could be wrong or sent by a malicious actor) will never solely determine a decision or process by itself.  Chainlink is currently partnered with Google, SWIFT, IC3, and Gartner among others. While these partnerships will provide tools (and traction), there is still the issue of attracting providers of primary source data.
The truth is what you make of it.
It does not take an article to establish that the truth is an elusive commodity these days. Trying to establish policy without access to verifiable baseline facts is not possible. Political gaslighting is a direct impediment to freedom. Stable coins and alternative mediums of exchange may get most of the attention, but keep an eye on the oracle solution. We may not all agree on what is “right”, but we should at least have a basis to determine that for ourselves.
 Commercial speech has less protection, defamatory statements are actionable and hate speech can be illegal.
 This is not an endorsement of perfect success in the goal of truth. There are many points where a malicious actor can attack, but in general, those are not actually within the fact validation and recording function, but rather the entry and exit points.
 For example, information from electrical, gas and water meters, can provide information relevant to the amount of manufacturing activity occurring in a geographical area (especially identifying trends).
 As with other typical distributed ledger paradigms, one can incentivize oracles towards good conduct and can rate oracles on the quality and reliability of the data that they provide (a reputation component).
 In some situations, “trusted” oracles may not have sufficient information to verify a “fact.” A system that provides degrees of confidence in the truth of a fact may be useful.
 The author has at times and may currently hold LINK (the native token of the Chainlink protocol).
 To the author’s knowledge none of these Chainlink partners collect primary source data (other than analytics of their own systems). While many of the major providers (Google, Facebook, Twitter) could take their data and attempt to be a data source, they all rely on their users to provide data (other than their own analytics) and therefore have the same issue that they have in guarding against fake news in the first place.
Also published on Medium.