Protect oneself from propaganda via fake books/articles

I think that there is a serious issue in how non-fiction books (as well as scientific articles) are distributed. Readers of books have no means to verify integrity of books. A book is a product, and each product has metrics of success: 1. count of distributed copies 2. amount of money earned 3. familiarity of population with the ideas within the book 4. acceptance of population the ideas within the book 5. recognisability of the title 6. reputation of the title 7. reputation of the author Authors usually optimize for 2, for them the rest of points are proxies for 6. Authors can say "if you want to be sure in integrity - buy it directly from the publishers we have a contract with". It is a working model, for example some software companies intentionally don't sign their software, to deprive owners of pirate copies the means to check integrity of them and so create more incentives to buy that software for $$$$$ from them. Imagine a book/paper having an undesirable effect for some powerful actor. Then that actor can spend money to intentionally disoptimize 3., 4., 6. and 7. of the original book/paper and its authors by making a fake book/paper having the same title but different, maybe even strategically slightly different, contents. And distribute it over resources hosting pirated content. Maybe distribute a fake paperback book in local shops too. Majority of books and media are consumed in a pirated form, and the consumers have never bought the original book/article to detect that "its" e-"copy" is fake. For example look at the picrel. The copy I have obtained has so many implausible inconsistencies (like a FSB operative routinely parking his car on the parking lot within the headquarters of Solntsevskaya OCG, or like gangsters praising the operative who has put them in prison), that makes an impression that the "author" is blatantly lying - exactly the effect a malicious nation-state may desire to achieve within the population: distrust to the author, to the title, and considering the title to be a taboo in any serious discussion. To solve this problem there is a need to check integrity of books against ground truth acquired from trusted sources, and the most importantly, wiuthout the collaboration of the original authors of books. Cryptographic hash functions are not helpers: a single text can be converted to many forms, each having a different hash. But possibly plagiarism detection algos can be useful.

>fake news >To solve this problem there is a need to check integrity of books against ground truth acquired from trusted sources Imagine falling for this hogwash

This isn't exactly what you are saying I know, but I like the idea of publishers releasing corrupted-but-correct-looking ebook versions on sites like libgen to make it hard to find actual pirate copies/ discourage people from going to the effort to pirate because it looks like its already available....but not the same book. You can generate books probably by telling chat gpt the title and blurb and getting it to pad out a plausible book and maybe inflate it to the correct ebook size with garbage

>This isn't exactly what you are saying I know, but I like the idea of publishers releasing corrupted-but-correct-looking ebook versions on sites like libgen to make it hard to find actual pirate copies/ discourage people from going to the effort to pirate because it looks like its already available....but not the same book. You can generate books probably by telling chat gpt the title and blurb and getting it to pad out a plausible book and maybe inflate it to the correct ebook size with garbage. You cannot do it without harming the reputation of the original author and the title. And it is a net loss in sells: majority of those who get pirated books wouldn't have bought them anyway. Imagine a dialogue >- I'm considering buying a book N of the author M >-Waste of money, I have read it, the author is spilling bullshit here and there. For example ... So it is likely that for the most of books the loss of sells originated from loss of reputation would outweigh the increase of sells via incentive creation. And the incentive is created not so from fake books, more from ambiguity in whether a certain copy of a book is genuine or fake. The only case when anybody is interested to do any activity about spreading fake books is to discredit the book and its authors and, maybe, censor the facts in genuine book (this is a secondary optional side-effect). What we need is a semi-centralized database of perceptive hashes of book contents, certified by librarian volunteers. A librarian registers on a website, the website does background check on the librarian, the librarian accepts that he understand the procedure of adding a hash (an original book received by library via official channels and only when there is no doubt in integrity of them is digitized (if not digital-native), and then hashed with a special cross-platform program that first extracts texts, and then does an NLP preproc pipeline, which is main purpose to get exactly the same text blob from different books of different formats) then the librarian gets capability to send redundant perceptive hashes of editions. If multiple librarians from different jurisdictions have sent exactly the same perceptive hashes for the same title, then the mapping is published and everybody can rely on it as on a trusted fingerprint. Possible attack vector is not only a Sibil attack on the website itself by fake librarians, but just identifying the librarians that take part in the site and sending them the fake copies. That not only creates fake records in the website, but also feeds people, who use physical libraries for reading, the disinformation. Ideally publishing hashes should be a job of original authors, not the public, and ideally hashes should be not in a central registry but on official websites or other places credibly connected to authors (university pages, GitHub repos and so on). But Litvinenko is dead and Feltshinsky and Goldfarb seem to be unconcerned about the book fate. Exactly that makes the hypothesis that the copy I got is actually just fictitious bullshit and not a result of an infoop plausible: if the book was been bullshit all along, it is against the interests of authors to confirm that. So now there is ambiguity. We cannot trust the facts in widely circulating copies, we cannot even expect that these copies are the actual book. It is also unplausible that the authors who wanted to maintain any reputation could had written it: the quality of the narrative is so unbelievably low, that if we expect the copy was genuine, then the plausible hypothesis is just hiring the cheapest ghostwriter one can find to make this bullshit and then publishing what he writes without even reading it.

I was thinking about this recently after I discovered by accident that some online versions of Marx's small works have been censored to remove references to something very specific. I only found out because I was reading a paper version at the same time as the online version and noticed the difference.

>>26713 Totally agree. There was no "Uhhh... I'm gay! Gimme money!" in original Marx papers

>>26558 Funnily enough, some lainanons here claim that they actively spread subtly changed content for the fun of it (see zine). I'm certain the number of actors that do this is quite small, but its a decently serious hazard and capable of doing outsized damage to the future.

Develop infrastructure to decide on an authoritative digital copy. Have a network of people from different countries verify the digital copy against locally acquired physical copies and use that authoritative digital copy to verify other digital copies. Create a service where a user can submit their digital copy to your team to verify it for a fee paid in crypto which will fund the infrastructure. These verified digital copies can be distributed with your cryptographic signature of approval. People who like physical copies could check their physical copy against a signed digital copy. Individual level trust in the centralized authority can be built by checking trusted physical books against signed digital copies.

Every centralized org of that kind will be either subjugated or physically exterminated, jointly by all the actors it counterihnteracts to, it is a shared interest of all those actors. And those who acquire copies from official sources they consider as "original authors" or their authorized distributors - they have already paid to the distributors, why do you expect they are going to pay to anyone else? And if you target pirates ... do you expect them to pay at all? Your business plan is bullshit.