Meanwhile, the social media giants were too busy silencing and shadowbanning foes of jihad violence and Sharia oppression to take much notice.
“Isis propaganda library detailing how to carry out terror attacks discovered online,” by Rhiannon Williams, inews, September 4, 2020:
The cache was found in the wake of the death of so-called Islamic State leader Abu Bakr al-Baghdadi, purported to be pictured here in a militant video (Photo: Uncredited/AP/Militant video)
A gigantic library of so-called Islamic State propaganda has been discovered online, some of which details how to carry out attacks in what has been branded a “blueprint for terrorism”.
The cache contains more than 90,000 documents, videos and pictures and attracts around 10,000 visitors each month, according to the counter-terrorism think tank Institute of Strategic Dialogue (ISD).
The material includes powerpoint presentations and how to hijack planes, organise a coordinated terror attack and details of attacks including the May 2017 Manchester Arena bombing, London’s 7/7 bombing and 9/11.
Videos debating the merits of certain kinds of explosives, illustrated guides to concealing weapons and disabling surveillance and training materials on how to manufacture bombs were also found among the library.
“It’s a blueprint for terrorism, complete with footnotes” Moustafa Ayad, deputy director of the ISD told Wired magazine, which collaborated with the BBC on the report.
“It’s everything anyone with an inclination for violence would need to carry out an attack.”
However, much of the library also consisted of material depicting daily life under the regime, including school curricula, religious texts, and other pieces of propaganda….
As the data is spread across a decentralised system without a single computer server, it can easily be replicated and shared, making it incredibly difficult to remove the material from the internet entirely….
Clive Delmonte says
Perhaps those who have logged onto this material can tracked and their names added to watch lists.
elee says
Hear hear, This stuff will go on as long as their scripture explicitly commands them to go in for this stuff.
gravenimage says
Islamic State propaganda library detailing how to coordinate jihad massacres discovered online
………….
Hideous, but no surprise.
Mark Spahn (West Seneca, NY) says
https://inews.co.uk/news/technology/government-funded-ai-software-detects-94-terrorist-propaganda-uploaded-125840
“The Home Office has funded new software to automatically detect and block jihadist content from websites.
It has been trained on more than 1,000 pieces of Daesh footage, and detects 94 per cent of terrorist content with 99.99 per cent accuracy, the Home Office said.”
What does that last part even mean? What is 94% of what (number of characters? words? sentences? paragraphs?) and what is 99.99% of what?
And why does this whiz-bang software not have a snappy name?
b.a. freeman says
it could be called AI, Mark, although it’s most likely just a bayesian filter that looks for material similar to that with which it was trained (the 1000 documents), and it probably operates much like any garden-variety email spam filter. the software is written so as to treat *ALL* the input/training documents as *definitely* poisonous, almost certainly with another batch of documents, which batch should be close in number to the batch of poisonous ones, that are verified as innocuous; it matches phrases, sentences, and words from incoming documents against the verified poisonous *and* the verified innocuous documents. after the filter has been started up, the user should look through both his “poison” and “innocuous” folders to mark false positives (innocuous documents erroneously matched as poisonous by the filter) and false negatives (poisonous documents erroneously marked as innocuous), all of which add training to the filter. as more and more documents are checked, the filter gets better and better, while false positives and negatives get more and more rare.
of course, if U never check your spam folder, then U’ll end up with no corrections, and also with lots of false positives (innocuous marked as poisonous) and false negatives (poisonous marked as innocuous), and the filter will *NOT* improve. i haven’t worked directly on software implementing bayesian filters for a couple of decades, but there are a ton of open-source filters readily available for adaptation (i.e., the coder doesn’t have to modify the filter; he just has to write a framework that implements the window and its controls, etc., that calls the filter), not to mention a ton of proprietary implementations. the HO had a ton of choices, and probably didn’t have to write a lot of software from scratch (few programmers bootstrap their code from the ground up, anyway), so it’s probably pretty good. bayesian statistics have been around since 1763, so i think that there’s a pretty good chance that the kinks in the math have been worked out.
none of this is to say that bayesian statistics cannot be misused. as (mark twain?) once noted, there are lies, damned lies, and statistics (personally, i might include nazicrat lies just above statistics). this witticism perfectly illustrates the fact that if sampling is done wrong (in a plethora of ways, whether intended or not), if data is “cleaned” (extreme outliers removed from the data) improperly, and so forth, entirely incorrect (and in the case of the nazicrats, entirely intentional) conclusions can easily be obtained. nazicrats, besides using brownshirt tactics (rioting in the streets, physical attacks and murders of opponents, silencing opponents by drowning out speakers, etc.), ask slanted questions in their opinion surveys (e.g., “answer yes or no: have U stopped beating your wife?” – although a simplistic question, illustrates how an on-line questionnaire which only allows certain answers can deliberately skew data). they can write bayesian filters that remove answers that they don’t like (supposed removal of outlier data). they can use re-sampling to enhance the effects of data they do like. there are innumerable ways to cheat, and the f***ing hard-core leftist liars in the legacy media/nazi propaganda outlets use them all. the bottom line is that U should not trust any survey whose complete unfiltered dataset and complete procedure used to arrive at their conclusions are not available to U. and, the HO being the HO, i have no doubt that the software they wrote will soon be trained to arrest brexit supports, christians, hindus, et al., especially if they post on fascistBook, twister, instaCrap, or any of the other nazi social media.
b.a. freeman says
“It’s really all about privacy, freedom and encryption. That’s what attracts jihadists.”
—
this final sentence of the article is grimly humorous; the pious ummah, which if fighting tooth and nail to end freedom everywhere, *depends* upon freedom and technology developed by free people to attack freedom. if ever there was evidence that islam is a parasitic brain disease, this is it.