There's Probably Nothing We Can Do About This Awful Deepfake Porn Problem

we built a global instantaneous communications network, and we're going to keep on living with the consequences

I recently was on the podcast What Could Go Right?, which you can check out here.

One of the very first posts I did for this newsletter was an argument that we can’t (as in, are unable to in real-world terms) censor far-right content online because of the basic reality of modern communications technology. The internet makes the transmission of information, no matter how ugly or shocking or secret, functionally impossible to stop. Digital infrastructure is spread out across the globe, including in regimes that do not play ball with American legal or corporate mandates, and there’s plenty of server racks out there in the world buzzing along that are inaccessible to even the most dedicated hall monitors. You can exclude certain people or certain messages or certain types of content from certain networks. But we have no ability to simply prevent the transmission of bits that someone really want to transmit, and in that piece I named a lot of examples, as well as showing that even in the pre-internet era actually stamping out extremist ideas proved to be impossible. To pick a salient example, Alex Jones has reached a huge audience and raked in tens of millions of dollars since he was supposedly “deplatformed” years ago. Because you can’t stop the spread of communication in the 21st century, even when you really, really want to.

What’s funny is how many times that piece has been referred to angrily by people who simply refuse to accurately understand its thesis. So, so many responses to that piece have said “look, Freddie deBoer says we shouldn’t censor Nazis!” I have in fact been named in some of the endless number of (subscription-harvesting) anti-Substack pieces in that way, as an example of someone who thinks we shouldn’t censor extremist content. Well, it happens that I am one of those free speech absolutists, yes, but that is very explicitly not what the piece argues - it’s precisely an argument that whether we should censor is entirely moot, because we can’t. The technological impediments to cutting off the flow of information (at least that which is not tightly controlled at the supply-side) are now existential. The establishment governments of the world were unable to stop ISIS from effortlessly spreading their propaganda at the height of the international campaign to stop them; I promise, you can’t stop people from spreading (say) pseudoscientific racism online. This is a reality people have to accept, even if - especially if - they think that reality is corrosive and ugly. I suspect it’s a similar story with all of this horrible AI “deepfake” celebrity porn. It’s awful, we should absolutely work to keep it off major networks and platforms to whatever extent we can, and it’s likely literally impossible to stop it from spreading entirely, given the structure of the modern internet. It’s terrible, but this is the world we’ve made.

The trouble is that, as I’ve seen again and again, in this era of entitlement people think saying “we can’t do this” necessarily means “I don’t want to.”