 "[The moderator crackdown is] just a reminder that anything you post on any of these platforms can and will be used for profit. It's just a matter of time until all your messages on Discord, Twitter etc. are scraped, fed into a model and sold back to you." β @ben h/t @dangillmor 41 comments
 "The only winning move is to refuse to play!" - Stop using, feeding or otherwise contributing to said platforms outside of being a paid employee for them - anything else is just oneself willingly donating time and labour to a corporation! @Woodswalked @molly0xfff @ben @dangillmor Which they legally can't because I refuse to sign off my authorship. ^  Best is to actually read ToS before signing up to a site... I had aleady 15 years ago decided to not join specific forums, that i should socially and technically have been pat of, because the knowledgeable and social people are using then But they had in their ToS that every writing, every photo, gets perpetually licensed to the owners of the forum "and partners". Or in plainer words, already back then they said they will use my data for all they want @molly0xfff@hachyderm.io @ben@m.benui.ca @dangillmor@mastodon.social Yah it took several days because they only let you delete 5 unaccepted answers a day, but I finished this morning. @shom @molly0xfff @ben @dangillmor Ever heard of low-background steel? Various instruments are made out of world war 2 and earlier battleship steel, some recovered from the bottom of the ocean, because all post-1945 steel is radioactive. Pre-2023 Internet content is going to be a precious resource like low-background steel someday, the way things are going. Archives will be valuable, especially if one has the legal right to use them. (Although that will NOT deter people, they will obfuscate.)  @mike805 that's a great analogy. archive.org has always been an invaluable resource, it's a real treasure now.  Low-background steel is only needed in a few specific applications, like radiation detectors. For everything else, regular post-nuclear steel is fine. βAIβ slop, on the other hand, is good for almost nothing.  @molly0xfff @ben @dangillmor now that we have open alternative it's time to ditch those people just to show them again that even a powerfull company is nothing without user #Twiitter @molly0xfff @ben @dangillmor who else is bored with the reddit narrative in 2024 It's kind of ironic they ban using generative AI to post on Stack Overflow too. They don't want their data getting tainted I guess https://meta.stackoverflow.com/questions/421831/policy-generative-ai-e-g-chatgpt-is-banned?cb=1 @molly0xfff @ben @dangillmor this is doing great damage to current web forums and any that get built in the future. It's going to make people less inclined to take time and help, or even post incorrect answers to mess with the AI. @BenjaminKlein @molly0xfff @ben @dangillmor Not saying this isn't grim or that it won't have consequences, but I wonder how many people really care about their posts being gobbled up and sold back to them. I mean, naively, one would think that millions of people would have problems with phones monitoring their every internet activity and purchase and tracking their every move, but here we are, in a world of corporate surveillance that makes the Staatsicherheit look like amateur hour. @molly0xfff @ben @dangillmor @molly0xfff @ben @dangillmor what's the deal with websites with "stack" in their name being pieces of shit these days?  @molly0xfff @ben @dangillmor Question: "scraped, fed into a model." Can't people scrape posts, etc. from Mastodon? They're not exactly hidden from the Internet... (Seriously, I don't know how this works) You're right, they can, but not without committing copyright infringement. But in the terms of service on sites like Stack Overflow, you usually give them permission to feed your content into βAIβ. @molly0xfff @ben @dangillmor canβt be repeated often enough: "[The moderator crackdown is] just a reminder that anything you post on any of these platforms can and will be used for profit. It's just a matter of time until all your messages on Discord, Twitter etc. are scraped, fed into a model and sold back to you." @molly0xfff @ben @dangillmor In a future of static information, turncoat programmers play cat and mouse games against the Resistance to harvest the underground forests of mycelial knowledge. Resistance programmers undergo rhizomatic dna changes to enable their fingertips to enter the fungal network.  @molly0xfff @ben @dangillmor The greed mongers have to even profit off of thoughts, nothing is sacred or private to them. Who cares? I mean, Google and the search engines have made money on SO already for the results just by ads alone. Why the stance now? To be clear.... any public data is monetized....Don't fool yourself. Federation (public) can be pulled and used for monetization too, because it is open by design. @bennysp @molly0xfff @ben @dangillmor many were under the impression that our content on StackExchange is licensed under CC-BY-SA which requires attribution. Reliable attribution is not possible in LLMs, as opposed to search engines @mattsains @molly0xfff @ben @dangillmor But when you search Google and find your answer on SO at the top, what made anyone think this wasnβt being monetized? @mattsains @molly0xfff @ben @dangillmor I will go further... what makes anyone think offering an open license model solution that is public, will prevent anyone from finding a loophole to monetizing? @mattsains @bennysp @molly0xfff @ben @dangillmor Apparently CC as an organization thinks AI training is fair use? @bennysp @molly0xfff @ben @dangillmor it can be, but that is also a choice, we made a long time ago under very different circumstances, so very much worth reconsidering now we have a clearer sense of how this might play out. Or you know we could have paid attention to the enclosure actsβ¦ Time to switch to Codidact. https://software.codidact.com/ It's the same thing as StackOverflow, only run by a not-for-profit. @molly0xfff @ben @dangillmor AI is going to ruin the internet. Everyone will be alone, an island unto themselves, surrounded by a sea of bots. And if we see another island, we will have no way to know if it's a bot on that island. The internet has stopped being real. @molly0xfff @ben @dangillmor I honestly wonder where these companies think future information is going to come from once they inundate all our current collaborative discussion forums with computer-generated regurgitations You might see it fom this perspective: These companies might not be interested in future "Information", but just in "something that users use" Thats why A"I" LLM are so popular right now, it does not matter how inaccurate or even plain wrong the "Information" spread by them is, it just matters that users use it  @molly0xfff well they sold the data for a chunk and open ai wants their moneys worth. The. Product has no say  @molly0xfff @ben @dangillmor The funny thing is that their very own rules require linking to the answer and the author for any derived works. Have they changed that recently/silently?  @molly0xfff @ben @dangillmor Welp (cracks knuckles) time to post some bullshit on SO and approve badly incorrect answers. @molly0xfff @ben @dangillmor That's just another reason, folks, that it is a moral imperative to use online platforms exclusively to share photos of fat, throbbing, transexual cocks with your mates. Train you AI models on this for size, billonaires! @molly0xfff @ben @dangillmor we thought Web 3 was going to be a crypto con, but itβs actually a hallucinated cesspool of AI generated content. |
Gregory's Server
@molly0xfff @ben @dangillmor
Then defo DON'T implement a #fediverse #ToS or #UPPβ’π«π€π©π β policy server or per user that stopped #Xai's #grok at #Xitter and on infoseceXchange. Don't. No. Stop. Don't.
Definitely DON'T have your own #UserPrivacyPolicy on each of your services. No. Stop. Don't.
https://infosec.exchange/@infosec_jcp/111524709239145200