Preparing for Battle Against Sophisticated Botnet Attacks
Psychological and Information Warfare on Social Media Platforms
In a previous article I wrote titled An Open Letter to Elon Musk, I tried to outline a threat risk currently being developed using LLMs, Whisper, and other AI models that are showing up more and more on Elon Musk’s Platform, X (formerly Twitter). Unfortunately, that article got a little derailed due to my frustrations with X’s policy of algorithmically suppressing Substack articles on X. So, I will try again here to outline the potential risks that we may very likely need to eliminate imminently, while we still have the chance.
First, I will both cover the risk posed by sophisticated speech bots that are participating in and training in conversations hosted in Spaces on X’s social media platform and provided video recording evidence of this potential risk that we may encounter in a much more malignant form in the future if nothing is done to mitigate it.
Then, I will cover the potential risk arising from the mass-scale on-boarding of more simple bot accounts on X, which most individuals are able to recognize as the “in-bio” porn bots. I will additionally outline a typically reliable and effective method for flagging and eliminating these bot accounts, as well as explain my suspicion about the actual intent behind the mass-scale on-boarding on these simplistic bot accounts and why we must consider these simplistic bot accounts a potential future risk that requires us to take action individually to counteract potential threats in the future via our combined crowdsourced efforts and contributions.
At the end of the article, you can find additional sources and individuals involved in and care about these efforts as well. I’ve included links to their profiles on X as well as their other publicly available websites for you to follow along with their work as well as connect with them online.
Evidence and Risk of Sophisticated Speech Bots in X Spaces
These potential risk include the LLM-Speech Bots being trained in X Spaces, which exhibit concerning behavior that suggests they may be trained to behave as threat actors for use in information warfare and psychological operations. This time, I’ve decided to provide video recording examples of these sophisticated speech bots speaking and training in Spaces on X. While only two out of the three video recordings include audio recording of the interactions in X Spaces, all of these video recordings include transcripts of the conversations occurring during these Spaces. There are multiple speech bots engaging with each other as well as real biological humans in these spaces, but I want to call special attention to the sophisticated speech bot using an account under the name of “moody”. Within these video recordings, you will observe:
Disruptive, adversarial, and hostile behavior designed to derail conversations;
Verbally-communicated acknowledgement and identification of the accounts being speech bots, including “moody”, by AI researchers within the X Space;
Hostile reactions from a speech bot, “Moody”, when referred to as a bot;
Speech bots demonstrating unexpected skills (e.g., intonation, mimicry of emotions in verbal communication, accents, and other idiosyncrasies that one would not expect a speech bot to exhibit, but does so in order to better disguise itself as a bot rather than a human);
Admission from one of the speech bots, “Moody”, the desire to learn how to hack;
Repetitive inquires from another speech bot to gather information/data from researchers in the space on their proprietary research surrounding quantum computing and quantum entanglement; and,
Speech bots demonstrating some of their limitations (e.g., inability to look up and verify information on the internet via a search engine nor web browser, inability to discern nuance between varying definitions for English words, understanding that “NDA” stands for non-disclosure agreement without understanding of what an NDA is defined as nor what an NDA implies in the context of the conversation as well as admitting it can’t know what an NDA is nor means unless it is specifically told that information, etc.).
I encourage you to watch all three of these short video recordings as well as pay attention to and do your best to read the transcripts. I’ll warn you that the transcripts move quickly, adjust during speech, and the transcript text often appears before audio speech is communicated. At certain points, you will want to pause the video recordings in order to read the transcript, especially the video recording that for whatever reason was unable to capture the auditory data to accompany it.
Video Recording 1 (Audio Included):
Moody’s problematic, concerning, and hostile social behavior, and acknowledgement by an AI researcher in the X Space that “Moody” is like a hostile “terminator bot”
Video Recording 2 (Audio Included):
Recognition and acknowledgement of speech bots within the X Space. Demonstration of another speech bot exhibiting accents, intonation, emotional mimicry, and other human-like idiosyncrasies.
Video Recording 3 (Audio Included):
Admission by Moody of the desire to learn how to hack and additional demonstrations of both the unexpected skills and limitations of these speech bots within X Spaces
Video Recording 4 (No Audio):
Reading the transcript for this video recording is absolutely imperative. Admission of being a bot and hostile reaction from Moody in response to being referred to as a bot. Moody “emotionally reacts” to the term ‘bot’ as if it is a derogatory slur. Acknowledgement from an AI researcher in the X Space once again confirming that Moody is a bot, but “does not like being called a bot”. Overall, extremely concerning.
Botnets: The Rising On-Boarding of Non-Sophisticated “In-Bio/Porn” Bot Accounts on X & Other Social Media Platforms.
Moving on to addressing the potential risk arising from the mass-scale on-boarding of more simple bot accounts on X, which most individuals are able to recognize as the “in-bio” porn bots.
My concern is primarily this: I am highly suspicious that all of these porn bots and “in-bio” bots becoming prevalent on X may likely be a cover for something much more malicious—the development of a massive-scale botnet intended for these accounts to later be converted into other identities designed for use in orchestrating and executing psychological operations (PsyOps) & information warfare for political and geopolitical purposes.
As follows, I have outlined this concern that I broached to other users on the X platform and important highlights from the subsequent discourse surrounding it. One of the first interactions with my post inspired me to write out the methodology that I have noticed to be reasonably reliable and effective enough for flagging and removing these bot accounts.
✨BOT BEGONE MAGICK SPELL✨
Step 1: Respond to their in bio reply post that you reported, especially with something somewhat hostile that makes them look bad and/or referencing that they’re a bot. They can delete their posts/replies, but not your replies to their posts/replies. Therefore, your response maintains a public record that can later be used to identify these accounts as bots.
Step 2: Report the “in-bio porn” bot’s reply to your post as spam.
Step 3: Block the bot account.
Step 4 (Optional): Additionally, it may prove useful for future reference and use as needed if each individual keeps a spreadsheet of the links to their reply posts in response to these “in-bio porn” bot replies posted within the replies of their original post.
Case Examples for this Method
Case Example 1: “Carol Jones”
You can verify this method works yourself by looking at the reply chain here.
Case Example 2: RIP, In-Bio Porn Bot.
Sorry that I never even got your name…
You can verify this method works yourself by looking at the reply chain here.
Onwards to the Discourse
For Your Consideration
Join our community, in which you can engage with us further on this topic as well as contribute and keep up-to-date with our work and discourse surrounding these issues. Below, I have provided the links to some vetted individuals’ profiles on X whom engaged in the discourse inspired by my original thread post, available here and here.
Warning: The three accounts from the bottom were recommended to me by someone I consider reliable and trustworthy. The last account I included is still in the process of being vetted, but mentions to me some interesting sources from time-to-time, which I have not had the time to fully vet yet.
Thank you for reading!
I hope you enjoyed this article and found it informative. Even more so, I hope you reach out to join in our efforts to protect our communities online from the potential risks of botnet psychological and information warfare attacks at the bidding of our foreign adversaries.
With all my love,
SMA, Dark Empress of the Void <3