http://www.scpwiki.com/antimemetics-division-hub
Imagine fighting something that actively eats any knowledge about itself and anything you know about it, including yourself.
Favorite pun: "digitized" to describe being absorbed by a monster consisting of a mass of fingers...
A lot of people have a bad reaction to http://scp-wiki.wikidot.com/scp-2718 . An entire "canon" (collection of stories and sometimes some SCP entries) that can be similarly rough is http://www.scpwiki.com/end-of-death-hub . Here's a roughly-equivalent reddit question: https://www.reddit.com/r/SCP/comments/ai3aik/recommendations... & if you're serious, by all means feel free to post your question there, the community as a whole never tires of answering it (no sarcasm).
But I would just say in general that if you've got a "hole" in your psyche, hundreds of writers across thousands of articles will eventually find it. No joke. For instance, I'm sure a lot of people would just bounce off of http://www.scpwiki.com/scp-5045 but it spoke to me, probably because I've played rather too many video games from an era where those sorts of graphics were current and I've got neural pathways adding overtones related to the graphics that a younger person probably doesn't. (But said younger person might have found Herobrine stories legitimately spooky at some point.)
(5045 is a bit of a "format screw", which is to say, it has stuff hidden beyond the usual text, images, and clearly-labelled expandable drop-downs that most of the Wiki is built with. Be sure to click on the images as you go along. Most images in the Wiki don't do anything but many of these do. But not all.)
I felt a peculiar chill down my spine the day I learned about Roko's Basilisk and Information Hazards: there must exist some thought/concept/idea that just learning about it puts your life in danger.
"[T]he founder of LessWrong, Eliezer Yudkowsky [...] reacted with horror. [...] Yudkowsky said that Roko had already given nightmares to several LessWrong users and had brought them to the point of breakdown. Yudkowsky ended up deleting the thread completely, thus assuring that Roko’s Basilisk would become the stuff of legend. It was a thought experiment so dangerous that merely thinking about it was hazardous not only to your mental health, but to your very fate."
https://slate.com/technology/2014/07/rokos-basilisk-the-most...
To be honest I'm not sure I understand this one exactly, but given the definition of <information hazard> I'm not sure I want to go deeper than this comment ever again.
There are known real informational hazards for humans, but they generally only affect people with epilepsy or those predisposed to staring at odd optical illusions for longer periods of time than most would consider reasonable.
The most common informational hazards are varying forms of behavioral modifiers, e.g. advertising, religions/cults, state-sponsored propaganda. Edward Bernays' is one of the more interesting people to study here, both through his written works (namely Crystalizing Public Opinion and Propaganda, both from the 1920's) and his later actions (popularizing female smoking, helping fruit companies and the CIA overthrow South American governments, etc). Roko's Basilisk falls into this category, if you want to think about it really hard it can hurt you but(somewhat ironically) the only people really susceptible are pseudo-intellectuals who congregate on places like lesswrong.
Next up are simple forms of adversarial inputs such as discordant sounds or rapidly blinking lights. People's susceptibility to these varies, generally only those with neurological disorders (epilepsy, severe misophonia) are severely impacted by known tech. Applied examples include LRADs, dazzlers[0], and/or some bowling alleys.
Last up there is at least one true informational hazard where perception alone is enough to disrupt the underlying structure of your brain (or at least your visual cortex). It's called the McCollough effect, and I'm not going to link to it but if you're curious a quick glance is generally safe. On the surface it's a simple color shifting optical illusion, but unlike most of those, if you stare at it for more than a handful of seconds, the aftereffect lasts for months.
[0]: https://en.wikipedia.org/wiki/Dazzler_(weapon), the boring ones are based on just blinding while the interesting ones attempt to disorient your brain with high frequency color patterns
I also love the adversarial inputs and McCollough effect. Another example of what I mean: it shows how we're fundamentally just machines that have exploits like any other, and in a very dangerous way.
Thanks for sharing; I find stories and topics that try to elevate and glorify mankind tired and repetitive. In contrast I find things that reveal the dark sides and weaknesses of humankind endlessly fascinating. Unfortunately those topics don't seem to be as popular to create content about.
This study[0] discusses adversarial images which fool both ML classifiers and time limited humans, a rather interesting development in the field.
It's a good space to keep an eye on in general, at some point someone's probably going to discover a snowcrash-level informational hazard and it would be nice to have some warning.