Podcast | What is Infolklore?

The Poison and the Antidote

The poison was never hard to find. The poison finds you. Turn on anything russian and it arrives by the tanker load, free of charge, around the clock. What took months to find was the other thing: the antidote.

Here is what I published on the 22nd of July this year, in a Substack piece called It’s Amazing What Regular People Can Do. One challenge sits under everything: we are swimming in Disinfolklore. Infolklore is scarce.

Two coined words in that quote. Disinfolklore is my name for folklore-like information told with meanness — stories built to bend our moods and our intentions, our attitudes and emotions. Infolklore is the positive kind: true stories, told with generosity, told by someone with real stakes in a real problem.

The first kind, Disinfolklore, is an ocean, and we all swim in it every day, and it’s what motivates all my work. Because we swim in this ocean, I decided to try and work out what this ocean consists of. The second kind turned out to be scarce, almost to vanishing. This episode and this mini-series is the story of the search for it — the search for Infolklore. Months and months and months of failure. The best-loved books in the English language going into a machine, one after another, and coming back stamped: no.

Automating Disinfolklore

And for those who haven’t been listening: what I’ve been working on since January is actualising my long-held plan, which is to automate Disinfolklore detection — and, in our detection, to create an automatic engine that will enable us, in times to come, to plug in a module the way today we may surf the web or look at ChatGPT. In the future we will have cognitive modules. And what I’m trying to build is for the… So that when we are confronted with the vast amounts of information, and information flows that continue to increase, we will have a means of interpreting it really quickly — as quickly as we really need to. And so I call that cryptotypic engineering, or Automating Disinfolklore.

So, my attempts to build this machine. I started off with the 10,000 propaganda items that I collected and hand-coded in Luhansk, which I talk about a lot, back in 2015. And so for much of the year I’ve been teaching a machine to tell the two kinds apart.

Where the search ended was really not in any library on earth. And I went through so many texts and types of narrative, from EUvsDisinfo to war college reports to all kinds of different narrative forms. And the search ended, really, in th