The Crisis of Democratized Truth

Social media, the first amendment, a failed coup, and a solution

** Since publishing this post, I’ve had a lot of great conversations and have updated the recommendations to better reflect the collective wisdom on this subject as well as a roadmap to break such a large undertaking into focused and completable phases. Thanks! **

Charlie’s having a tough time

I have an almost 3 year old little girl who I, like every parent, struggle to teach about how the world works. Something I’ve always tried to push against is the temptation to over-use the word “dangerous.” For a little kid who doesn’t know how to be careful and when, just about everything is dangerous, but teaching a child that the things around them are stained with danger teaches them to navigate the world in fear.

So I try to use the word “powerful” instead. Because when something is powerful, you need to learn about how to use that power safely and how it can easily hurt you without safety measures and practice. Knives are powerful, fire is powerful…

Social media is powerful.

I’m going to skip the part where I try to explain how and why social media is powerful and how it’s tearing us apart from the inside out because I think we are all starting to agree about that. What I am going to talk about is exactly how and why we need create the desperately needed safety measures and a big Hail Mary of a solution that might help heal the fractures in our sense of reality and truth caused by the power of social media.

What happened on 1/6/21 in DC, my home town, shook me to my core as I’m sure it did to almost everyone else. It was both completely unsurprising and utterly devastating. Though the deplorable events that happened that day are bad enough, what weighs even more heavily on my heart is the fear of our ability to respond. The reason being that if we cannot clearly and collective respond with consequences that match the magnitude of the actions, then we further reinforce the fact that very little truly matters to us as a people.

The Black Lives Matter movement is a perfect example of this. The lives of black, brown, and indigenous people do, in fact, matter. They matter to themselves, they matter to those of us that love them, they matter to their Creator. But when something or someone truly matters, their pain, suffering, endangerment, disenfranchisement, and persecution require equal and opposite reactions of restoration, healing, and justice by those to whom they matter. So in this example, if we as a society are incapable or unwilling to effectively rectify the injustices inflicted upon black lives which it often seems that we are, then as heartbreaking as it is to say, black lives in fact do not matter to us as society.

So why am I talking about this? Because we desperately need to respond with action to things that matter, and I am terrified that we will continue in our inability to do so. So in the face of the innumerable reasons why the insurgency and failed coup in DC actually happened (which is only one of the many crises we face), fixing the problems caused by the unfettered power of social media might be the thing we have a shred of a chance at accomplishing. And by God do we all need a win right now.

A deductive approach to our deductive crisis

I am not a scholar on this matter in the least, but I do create tech companies and products for a living, so maybe that at least makes me a practitioner in the field. The approach I’m going to take is a bit like those of theoretical scientists, where they use the evidence gathered in experimental fields as their logical premises and deductive reasoning to create new conclusions and theories. These theories can be evaluated by the validity of their premises and internal logical integrity and eventually proven or disproven through experimentation. And ironically I’m going to be using deductive reasoning to talk about our cultural crisis of deductive reason — the way in which we use a confusing mix of information and opinion to deduce what is true about the world around us. So if my conclusions about how to solve this are wrong, you should be able to trace them back to the premises they’re based on. So let’s start there.

Premises

1. A conclusion is something that someone claims to be true, whether that claim is supported by valid premises and logic or not.
2. Evidence is an irrefutably valid premise that supports a conclusion, and can be something tangible like a photo or intangible like a quote attributed to someone.
3. Information for the purposes here are conclusions that are supported by a valid use of evidence, and thus should be collectively agreed upon as true.
4. Opinions are conclusions that are either not supported by what is agreed upon as valid evidence or the use of the evidence itself to support the conclusion is invalid.
5. Opinions are protected under the First Amendment but have been historically limited in their reach due to technological limitations (i.e inventions like the printing press and microphone) and the moderating duties of the press.
6. The purpose of the press as protected by the Constitution is to gather evidence and distribute information, thus being an agreed upon source of truth.
7. The press has traditionally been held accountable for attributing evidence to their conclusions by citing sources, publishing photos, etc.
8. The press have historically been held accountable by the government, the public, and their peers due to their privileged access to evidence with which they create information as well as their protected wide-reaching distribution channels that are used to disseminate information to the public.
9. Much of the press now runs on advertising models that are driven by levels of consumption, so many have resorted to adding more entertaining opinion mixed in with information.
10. Social media platforms have created new distribution channels to the entire population, and through those channels have empowered the people to distribute a mix of information and opinion in massive quantities.

Given the premises above, let’s move on to problems that we’re all experiencing that are the reasons for drawing some conclusions for action.

Deduced Problems

1. Both social media and the press have become a confusing mix of information and opinion and cannot be considered a trusted source truth, thus putting the burden of validating evidence on the individual. In the absence of doing that work, we reduce the truth to the level of opinion. (Let’s call this the classification problem)
2. Social media channels have surpassed the reach of the press and produce an exponentially growing proportion of available content, yet are allowed to self-regulate. (Let’s call this the virality problem)
3. There is no way to attribute supporting evidence to a conclusion at the scale of the digital age and have those relationships cross channels or audiences, making it logistically impossible to easilly identify anything as true. (Let’s call this the feasibility problem)
4. We now have a population that is sourcing their own truth via these democratized channels which is a duty that is constitutionally tasked to the press. (Let’s call this the channel problem)
5. Because the truth has been reduced to the level of opinion and we can pick and choose the sources that reinforce our personal conclusions, we now live in our individualized versions of reality. (No need to label this one, it just sucks).

So if you’re still with me and at least see the problems above as valid, then let’s take a swing at solving each and every one of them.

Proposed Solutions

1. We need to create a decentralized and highly secure database of primary sources of evidence. This would be something that would house public documents, photos, links to external sources, and quotes that have been validated by the attributed source. (Let’s call this the Evidence DB)
2. We need a technical infrastructure that enables people to link content and individual conclusions within content to the Evidence DB and empowers consumers to traverse the supporting evidence or lake thereof, thus solving the feasibility problem. (Let’s call this system the Attribution Infrastructure). Think of this as Wikipedia’s “References” section at the bottom of a page that would work for all digital content. It would likely take the form of a new URL/URI structure, much like how you can now create deep links to mobile apps rather than just websites.
3. We need to create a transparent non-governmental network of individuals and organizations with diverse backgrounds, perspectives, and beliefs that have the democratic power to 1: moderate and audit the information added the Evidence DB, and 2: moderate the linking of content to the Evidence DB and other sources, thus helping to solve the classification problem. (Let’s call this group the Moderation Network)
4. As digital content is often filled with conclusions of varying validity, a single piece of content would need to get a summarized score of whether it as a whole should be considered information or opinion. This score would likely be calculated based on the ratio of conclusions (let’s call this the Validity Score). This would also require a modification to the Open Graph Protocol so that score could be made available as part of the summary of a shared link just like a photo from a link can be displayed by using og:image .
5. Though all content creators, websites, and publishers should be able to use the Attribution Infrastructure to cite their sources, organizations certified as press should be required to do this. This press certification would be a non-governmental credential audited by the Moderation Network that would require certified institutions to classify all of their content and the conclusions inside as either information or opinion. The Moderation Network would have the power to revoke the press certification if they failed an audit too frequently. These would be the only channels protected by freedom of the press, helping to solve the channel problem
6. We need to create a threshold for audience size (let’s say 10,000 viewers and call this the Virality Gate) that restricts all unclassified content from further algorithmic distribution by things like “shares” and “retweets” until it is validated. Unclassified content as well as content classified as opinion would be limited to directly connected audiences, thus helping solve the virality problem.
7. When a piece of content hits the Virality Gate, the author would have the choice of submitting it and their citations for verification for a fee, which would be necessary short of massive government funding in order for this to be a sustainable model. If they choose not to pay for verification or the content was scored as opinion, the content would simply remain behind the Virality Gate, helping solve both the virality problem and the channel problem.
8. The channels that have directly connected audiences larger than the Virality Gate but are not certified as press should have all of their content classified as opinion and be protected as free speech but should also have the ability to submit attributed content as information to the Moderation Network for a fee, helping to solve the channel problem.

Something like this is obviously no simple undertaking and not something that could be accomplished all at once. So let’s try to break this down into a few phases that could create real change as soon as possible without having to wait for everything to be done.

Roadmap

Phase 1 — linking evidence

• We would begin designing and building the Attribution Infrastructure that would start by linking to other sources rather than the Evidence DB. This would allow publishers and content creators to begin citing their sources and make websites begin to function a bit like Wikipedia. This project would either need to be undertaken by a nonprofit and funded by donors (most likely route) or funded directly by the government (less likely for a starting place).
• In parallel, we would also begin empowering the existing community around Wikipedia to moderate the use of the Attribution Infrastructure as they are already highly engaged and could function as a first version of the Moderation Network. And if there were a paywall for something being verified as information, the operational structure could potentially become sustainable on its own.

The end goal of Phase 1 would be to complete the Attribution Infrastructure so that publishers, content creators, and readers could begin adding richer and more interactive citations as well as flagging statements made that need evidence to support them. This phase could be completed without any aid from government or cooperation from tech companies.

Phase 2 — building trust

• Transparency would need to begin to be built within the Moderation Network so that people could see who is a part of it, what actions they’ve taken, as well as nominate and/or vote for people to be added to or removed from it.
• Technical development of the Evidence DB could begin in parallel as well as the creation of the protocols needed standardize and secure how evidence is added. There would likely need to be an highly secure archival version that could only have data added and never updated, and an operational version that regularly copies the archival version and interacts with the Attribution Infrastructure.
• With hopefully increased confidence, early versions of the Validity Score could begin being displayed on websites with an interactive summary of the supporting evidence or lack-there-of that generate the score.
• Facebook would (hopefully) voluntarily update the Open Graph Protocol to accommodate for a summarized score through a parameter like og:validity_score that platforms could begin displaying as part of their social sharing summaries.

The end goal of Phase 2 is to build the structures of trust that will make or break the adoption and societal impact of a system like this. One way to think about this is that a representative democratic government is basically a decision system that enables a society to collectively decide what actions we will allow and not allow and what we will spend our collective money on. The gravity of a system through which we decide what is valid information and what is opinion I would say is on par with the former, and without it, our current sense of collective reality is similar to the geopolitical maturity of the Dark Ages — open to massive idealogical invasions rather than a more peaceful diversity of opinions and beliefs.

Phase 3 — passing regulations

• The first regulation that the government would have to pass would be to either allow the Moderation Network to credential organizations as protected members of the press or to at least give them the power to revoke or suspend the credential based on an organizations use of the Attribution Infrastructure.
• The second regulation would be to create the Virality Gate that would set a threshold for the viral, indirectly-connected distribution of unverified content.
• The third regulation would be the requirement to display the Validity Score on all shared content and potentially enforce the integration of the Attribution Infrastructure with the content inside of a social media platform rather than it existing only on external websites that posts can include links to.
• The fourth and final regulation would be to fund the operation of some or all of these structures. This could potentially function similar to the WHO as this is not just a domestic crisis, but a global humanitarian crisis.

The end goal of Phase 3 would be to make a societal commitment to the structures created in Phases 1 and 2. Without doing so, these structures could still exist, but would likely follow a similarly volatile and fractured path of other decentralized movements such as cryptocurrency.

So that’s it. A distributed moderation network, a massive technical infrastructure, and some laws about algorithms. To most of us, it would basically look like a subtle icon at the top of content signaling the classification as well as the ability to hover over some classified text to see links to supporting evidence — nothing we haven’t seen before.

If we have any chance of healing our fractured and tribal versions of reality, it is going to take a long time to get there and a lot of work to do it. But hopefully these conclusions turn something that may feel impossibly infinite in scale into something dauntingly huge yet possible. Our only choice if these problems truly matter to us as a society is to try something — anything — so why not this? And if you disagree, poke holes in the premises, clarify the problems, propose new solutions, and send them to your elected officials. Because by God do we all need a win right now.