The Plan

Any Online Driver of Massive Traffic can unwittingly become a vehicle for rapidly spreading poorly sourced information, or outright misinformation. There is an emerging consensus that the likes of Facebook and Google should do something more to prevent lies and silliness from becoming too popular online.

But let’s not reinvent the wheel.

Creating new authorities to somehow police the information online is impractical, worrisome, unlikely to succeed on a global scale, and it smacks of censorship.

There’s a better approach. Throughout the centuries, societies have nurtured and financed independent institutions entrusted with the task of researching nature and society – and getting one notch closer to the truth.

Those bodies are Universities and Academic Institutions. Academic research regularly fails, but it fails more gracefully than anybody else, and corrects itself. It’s their job.

Some people may not like specific Academic Institutions, and they should demand better vetting of their members. But nobody seriously thinks that we should get rid of Universities and Academic Institutions altogether. Their role in principle is accepted by time-honored tradition, and it’s not in itself a point of political contention anywhere. Academic Institutions also represent globally a plurality of points of view on just about anything where the matter is not settled. Their consensus, when it exists, might be wrong: but it’s still the best bet available.

Online sources affiliated with Academic Institutions (WAI for short) are the best existing bet on reliable, interesting information online. Because Academic Institutions generally vet their outgoing links, any online source directly linked from an Academic institution (DAI for short) is also a reasonable bet in terms of reliability.

The core of the 15mc goal is to have ODMTs increase the traffic driven to reliable online sources, namely the ones affiliated to Academia (WAI); and to online sources directly linked from Academia (DAI).

An example: bees

Let’s say the University of Alabama has a great site about bees, bee.ua.edu.

It’s popular, and people share it on Facebook. But if Facebook’s algorithm recognized it as a WAI (a website affiliated with an Academic Institution), the “value” of shares and likes could be optimized to make it appear in your timeline just a bit more often than a less interesting site. That’s good, because the University of Alabama created a very valuable, instructive, reliable resource about bees.

Nobody is going to censor or excessively penalize a site about bees that is poorly sourced or misleading – for example, BeesAreMammals.com

However, if our plan succeeds then the poorly sourced site BeesAreMammals.com will decrease in relative importance. That’s also good – because, quite positively, bees are not mammals. Society has an interest in not spreading such false information, just as it has an interest to not teach falsities about bees in schools. In short, we should just place a bit more prominently, in the global library of the Internet, the tomes that have good educational value.

A positive feedback loop

There are secondary effects, in this approach, that are just as important.

BeesAreMammals.com, the rogue website, probably wants to maintain its traffic and position online. If it as a for-profit site, it may therefore consider hiring an actual entomologist to provide valuable information. When it does so, and corrects the record, the University of Connecticut might be happy to link their renewed site WeWereWrongAboutBees.com

Now WeWereWrongAboutBees.com, linked by the University of Connecticut, is a DAI and can better compete for attention with the WAI of the University of Alabama, bee.ua.edu. The algorithms of ODMTs are considering both worthy sources, and both their traffic increases, relative to poorer resources.

Essentially, with this system Universities are given an implicit premium in saying what online sources are of high quality, and should therefore deserve more visibility online.

How big will the premium be? It depends, of course, on the algorithms of ODMTs.

15mc favors a gradual approach, where the premium is not very big at the beginning but it might be increased over time. Not only ODMTs will need to adapt, but Academic Institutions will also need to adjust to an increased role as online influencers. Tinkering is more likely to succeed than revolutionizing a complex system from day 1.

It is quite possible that even better positive loops may be generated. It is reasonable to expect, for example, that some Academic Institutions will be incentivized to attempt and produce popular and interesting sites, besides the technical ones, which is in line with their socially recognized overall mission of educating the public.  It is not unlikely, however, that something might go partially wrong in the process and might need further fix –  that’s why tinkering is a good idea.

How would it happen

Here’s a sketch of a possible scenario for the 15mc initiative to succeed.

  1. An ODMT that we’ll call “Z” accepts to be transparent about the percentage of their online traffic currently sent to Academic Institutions and sources linked from them – WAI and DAI. Purely hypothetically, let’s say that such combined percentage is today 13% of the outgoing traffic of “Z”.
  2. As a matter of self-regulation, “Z” sets a public goal of increasing the WAI+DAI percentage from 13% to 14%, in the course of 12 months. It may not seem much, but such a change for a major ODMT means shifting millions of global eyeballs from the likes of BeesAreMammals.com  to  the likes of bee.ua.edu (and WeWereWrongAboutBees.com). It may also instantly create a cottage industry of websites optimizing for actual quality and reliability, to gain traffic. It can shift the public sentiment  towards more critical thinking, which generates more positive feedback loops. It can foster policies and further steps in the right direction. And “Z”, of course will hopefully be emulated by other ODMTs.
  3. The public goal (13%->14%) will receive public attention and implicit independent monitoring. Some additional variables will need to be considered for the goal to produce desirable effects. For example, it will be a good idea to keep an eye on countries distribution and other factors.
  4. For “Z” to achieve their goal, a “Z” team of engineers working on “Z” ‘s algorithms and interfaces will figure the best internal technical strategy. The details of this technical implementation do not need to be made completely public by “Z”, because trade secrets are involved and there’s no reason to ask “Z” to act massively against their own commercial interest –  as long as it is working towards the positive publicly stated goals.
  5. Hopefully, “W” will follow “Z”. More ODMTs will join too. It will become at some point expected by the industry for ODMTs to have traffic goals that are in part measured by quality and reliability. There are countless possible ways to make the basic proposed system better and more sophisticated, but it’s important to start moving in this direction, with something that most of the public can understand enough and trust to be fair. Systems that are not based on something widely recognized as positively educational as  Academic Institutions at large, all risk to become divisive – and the ODMTs inevitably would risk to be deemed as partial in their choices, which is against their interest.
  6. When quality online improves, the public goals of ODMTs may be set more ambitiously for the subsequent years.
  7. The world will be a better place.

For a plan like this to happen, we need a movement of opinion in support of it.