"Same as it ever was"

If there's one lesson from the 2020 Elections it's the mass radicalization of a big chunk of Donald Trump's 74 million voters. A number of terrific books and articles have traced the process to the cycle of (fake) news, social media algorithms, polarization, and more fake news. These include the early work of Eli Pariser who coined the term Filter Bubble, Yochai Benkler's Network Propaganda, and more recently Sinan Aral's Hype Machine. A key article came from Michel Chollet in What Worries Me About AI. These and other analyses point out the speed, adaptability and scale of these new propaganda tools. But are they really so different from earlier communications tools? Are their effects uncontrollable or unmanageable?

Below is Chollet's concise chart on how social media algorithms target, polarize and control minds.

Looks scary - all these algorithms - how to control them?

If you look a bit more closely at the structure and process of network propaganda you see that the same process underlies all the developments of communications technologies once they establish enough scale and sophistication to be used as propaganda tools.

Local language printed bibles, books, pamphlets, newspapers, radio, film and television all went through similar periods of liberating discovery, broad adoption and eventual weaponization before societies developed norms and regulations to harness them, usually after wars and revolutions. These included the Reformation's Thirty Years' War that felled nearly a quarter of Central Europe's population, the Napoleonic Wars that devastated the same terrain a little more than a century later, and World Wars I and II which all told killed around 100 million people.

Political scientist David Stasavage, in The Decline and Rise of Democracy gives a good account of the effects of incessant pamphleteering on democracy in the 18th Century: 

"… Whig leaders …offered the justification that the Triennial Act, and the frequent elections it entailed, had stirred up public passions to such a degree that electioneering became a game of slander. It is true that even twenty-first-century purveyors of fake news would have been impressed by some of the vitriol that appeared in British print publications at this time...

For many in the American colonies, his case illustrated everything that was corrupt and tyrannical about the British parliamentary system."

His assessment points to an important omission in accounts of today's technological wave. My years in tech and media have shown me that no tool can creates thoughts or emotions that aren't already present in an audience, engrained and already gaining momentum, as Jacques Ellul described in his 1973 survey of Propaganda.

Even more important: the impact of propaganda is normally absorbed by dynamic social orders. It's when a regime or ruling order or ideology has past its point of cohesion, when it has atrophied as a belief system, that propaganda's asymmetric assaults on the establishment are terminal, toppling regimes. As social historian Eric Hoffer puts in in The True Believer:

"HG Wells remarks that at the time of the Reformation people 'objected not to the church’s power, but to its weaknesses'…

The French Revolution, which was also a nationalist movement, came as a reaction not against the vigorous tyranny of the Catholic Church and the ancient regime but against their weakness and ineffectuality. When people revolt ... they rise not against the wickedness of the regime but its weakness."

Today's network technologies are not so different than past disruptions. But what about their speed, adaptability and scale? Algorithms are super fast and flexible; social media platforms have billions of users. That is inarguably true. But those qualities alone don't give propaganda its asymmetric power: the power of a tool to topple an empire depends on the cat-and-mouse game between innovations, insurgents and establishments, and how quickly society adapts. Today's society is light years ahead of the staid orders that collapsed under previous propaganda assaults, like the French Ancient Regime, the pre-war 19th Century imperial order and the lumbering Russian Romanoffs.

The threat of persuasive technologies isn't diminished by placing it in the long history of communications innovations. Demystifying the black box software that platforms like Google and Facebook is part of reclaiming democracy itself. That in turn means using today's technological tools instead of yesterday's industrial solutions.

It means interoperability - the ability to see alternative choices and search facilities on dominant platforms like Facebook, Amazon, Youtube etc., so that one oligarch can't control the recommendation on his platform.

It means attribution - tagging posts and making their sources and their amplification (including who's paying and who's playing) clear.

Above all else, though, it means real-time peer-controlled mediation: instead of experts pondering from on high, its vital to use forums, juries and assemblies of everyday citizens to participate in disputes about news the way Wikipedia evolved its complex mediation systems.

But looking solely at media means dealing with the symptom, not the cause.

The real cause of today's democratic disillusion is the disconnectedness of urban, rural and suburban communities to the levers of democratic power itself. I've argued earlier that this can only be addressed by updating notions of property (to include your data and content), citizenship (without trying to shove everyone into a single set of norms) and nationhood (which is often too big to act as one and yet too small against platforms).

That will be the subject of a later post...