Asymmetric polarization, universal deepfakery

The Network Propaganda book by Benkler, Faris, and Roberts (BFR) continues to impress me. They provide a compelling counter-narrative to the "social media exacerbates polarization" narrative that I've been invested in throughout much of my recent work. This latter narrative gains a lot of traction from scholars like Cass Sunstein's #Republic: Divided Democracy in an Age of Social Media. Essentially, the neurocircuitry of us-versus-them that makes humans the biased, motivated reasoners we are plus the outrage/fear-driven attention economy of the current internet makes for an especially divided, polarized demos.

BFR's pushback (I'm just now getting to their chapter on polarization) seeks to undermine the "both sides do it" subtext of Sunstein et al.'s argument. Polarization, they insist, is asymmetric. Over the past generation, the USAmerican right has become much more rightwing than the left has become left. The left has its extreme actors, but on the whole there's nothing approaching the kind of debunked-but-persistent conspiracy theory narratives that regularly consume a vast amount of oxygen on the right. Whereas Sunstein (I'm generalizing here to mean the broad swath of folk arguing in his vein) would say that both left and right engage in identity-motivated, antagonistic communication, BFR's research points to a news media ecology with two fairly distinct sides: a basically truth-seeking, self-correcting center/left--and the extreme right. The right, they contend, has its own, long-in-the-making media sphere operating not according to verification protocols but to identity-affirming protocols.

BFR admit that this asymmetry is awkward for scholars who wish to preserve a sense of partisan neutrality. Rush Limbaugh has already branded academia one of the "Four Corners of Deceit" (along with government, science, and media). The idea that university scholars are left-leaning and hostile to the right (especially the Christian right) is a background assumption across the right. It is not without some truth; humanities and arts scholars do skew left. Most scientist accept anthropogenic global warming, evolution, and other right-wing bugaboos.

To say, as BFR do, that the right-wing media produces, signal-boosts, and circulates many more untrue things than the mainstream media--well, I can just imagine what conservative-leaning folk would say about that. How convenient. You find that your side is for truth while our side spouts propaganda. 

What's depressing is that, as BFR realize, the very model of network propaganda--the self-reinforcing ecology of right-wing disinformation and misinformation--inoculates itself against truth-checking by dismissing reality checkers as biased. PolitiFact? Snopes? That's just more corrupt, left-wing media obsessed with making Trump look bad. I mean, look how they find Trump uttering more lies than Democrats! Clear bias. And what if, in fact, Trump is spouting more lies--more by orders of magnitude--than any president prior to him? Please. 

BFR's case actually takes me a bit too far afield from the brief talk I'm giving tomorrow on deepfakes. Indeed, I don't think they talk about deepfakes at all. But their insights help me frame the field of dis/mis/information exchange in which deepfakes play.

Or do they? The fear of deepfakes, the thing that keeps security experts up at night, concern scenarios where deepfaked videos trick reactionary leaders like Trump into drastic actions. The reality of deepfakes, however, is that 96% of them are porn flicks with non-consenting women's faces superimposed onto actors' bodies. Beyond this considerable harm--one that doesn't fit BFR's definitions of mis- or dis-information--is the "worst thing about deepfakes" according to Thom Dunn: "we know about them." The accepted possibility that any video may be deepfaked grants those in power a handy excuse for those who would hold them accountable via video evidence.

I await the wave of deepfake defenses from authority figures like police ("that clip of me planting the drugs? Deepfake!"), politicians (deepfaked pee tape, anyone?), and celebrities  ("My ex deepfaked the footage of me hitting her!"). And they may be right! Deepfaked false allegations will soon be as easy as uploading a YouTube--easier still for those with power or platforms who already have access to scads of images/clips to train a deepfake app on.

Add facial recognition tech, and a whole vista of micro-targeted deepfake advertisements pop into being. See yourself--literally!--having fun at Disney World's newest attraction. Hear yourself, in your own voice, using your own style of speaking, pleading with yourself to buy X product, watch Y show, or vote for Z candidate. See your parents and closest friends pleading with you.

Or watch yourself committing a crime or deeply embarrassing act--caught on tape!--and learn that the video will be forwarded to everyone you know or work for unless you pay X amount in bitcoin or Libra.

I've switched gears mid-post. Awkward. But that's what happens when I'm writing a post and paper simultaneously.

More tomorrow, with a report on how the paper went, maybe.

JF