I find the concept of these 'fact checkers' a worrying parallel to the much more dangerous 'trusted flaggers' included in the EU Digital Services act. In both cases the ideas suffer from a host of problems. Some are inevitable consequences of handing power to the unaccountable, like bias, failure to consider context, overreach, inaccuracy and wilful misinterpretation. The other consequence is that such systems lend themselves to misuse.

In the case of your annoying facts being inconvenient for certain groups and individuals, being able to complain to Farcebook and have them pass the matter to a 'fact checking' organisation, is very attractive to people who cannot bear to lose face and are unable to admit they are fallible. Although it might be tempting to just shrug off being 'fact checked' and (incorrectly) found to be wanting, if I was in your position I'd be absolutely hopping mad! It's not just that your research is being casually disregarded and your information is wrongly tagged as untrustworthy; it also means that if a follower posts a link to your information, they are also tagged as posting untrustworthy or discredited information. That's a heavy disincentive for people to amplify your voice. A nasty payback for daring to speak out of turn, with full anonymity to the complainant and at no cost to them either.

The 'trusted flagger' idea has all of the flaws that the 'fact checker' has, but much, much higher stakes. A person who can influence the trusted flagger system to their advantage can unleash hell on someone if they want to. The kind of hell that starts at 4AM with the inevitable size nines kicking in your front door and dragging you, handcuffed, out of the house in nothing but your pyjamas. Whether it is through a bribed or blackmailed flagger, an individual flagger with a personal grudge or some other human factor, the 'trusted flagger' system is inescapably capricious.

Much worse is how the data collected by the flaggers and their associated data miners will be used. My worry is that this data would be the seed for a 'social credit' system. I'd hope that some time shortly after that, Orwell's speed of rotation in his grave would pass light speed and cause the spontaneous formation of a small black hole, erasing us from history.

Another worrying part of the 'trusted flagger' system is that the language used in describing one says: "Trusted flaggers are entities, explicitly not individuals, which must cumulatively fulfill the following conditions established by the legislator" The various conditions don't require that the trusted flagger is human. I think this language is deliberately written to allow using AI systems as 'trusted flaggers' in addition to the police and the host of other trusted organisations. There is also language which gives full access to the copyright mafia for enforcement. Chilling: https://www.lausen.com/en/the-trusted-flaggers-in-the-digital-services-act/

The near future looks more and more like an episode of Black Mirror.

Expand full comment

Comprehensive and persuasive in your usual thorough skewering approach. Well done, and as noted LinkedIn applied the dreaded "misinformation" censorship to my repost of your Spiked article.

Expand full comment

It's almost as if there's a theme, or thread, here. Perhaps the way that newspapers are becoming more and more opinionpapers the 'factcheckers' should really be called 'opinioncheckers'.

Expand full comment

Here king, you dropped this... 👑

Expand full comment