#mythbusting

it does not make sense to have "invisible magic discord bots which are hidden in your server" as a result of a supposed partnership between discord and some ai company, because if discord wanted to secretly give some ai company your messages or uploaded images, then

they don't need a fucking bot to do this, they already have the data

what makes you think you would able to counteract this by using discord's features

furthermore, "partnerships" are mutual: if a company claims that a partnership exists when no such partnership exists, discord can tell them to knock it off or announce that no such partnership exists, and guess what, the article does not mention any such partnership, it says "we launched an app", which is like, whatever, it's inside all the other shovelware hidden in the menus no one clicks.

the evidence for "invisible magic bots" is super flimsy and results in misunderstanding about how banning works (banning a user by id and then seeing the entry in audit logs). in reality, you can ban users that are not present in your server, which I've been doing as a preventive measure for years at this point when I see someone on another server doing stuff I would be banning for if they were present on my server.

you also can't add apps or bots to your server without having "manage server" permission.

furthermore, as a bot, you can't read messages on the server without opting in with enabling message content intent, and needing explicit approval from discord for this.

(this is what you see in discord developer portal in the bots settings. "Read more here" leads to this website)

in fact, "application commands" discord feature that the bot uses was introduced specifically so bots don't have to read every message on the server to see if it starts with an exclamation mark or whatever in order to know if to react to it, and how.

there are also "external apps" which are per-user, rather than per-app, but these are basically equivalent of a user DMing a bot, and they don't have access to anything aside from whatever user sends them, and, if the role has has "use external apps" permission enabled in the channel, they can send a message as a result of a user action.

and yes, potentially a member of your server could send some image in it to an AI company by using the application command, but then again they can also right click -> save and upload it to chatgpt. your threat model is fucking bonkers.

guest at the museum: oh look, a fainting couch!

me, preparing to talk to him: :D

him: it's not surprising- I mean, if you look at some of those corsets-

✦ On Shiny Things & Structure ✦ A skeptical reflection on hypnosis trends. It dismantles the allure of NLP and redefines the true requirements for profound trance. Consider this a philosophical fragment for practitioners.

✦ ᚺ ᚢ ᚲ ᚺ ✦

Then comes NLP. Something modern. Something shiny.

Yes—thought and language matter. Yes—meaning shapes experience.

But if Neuro-Linguistic Programming were the holy grail, the pot of gold at the end of the rainbow, wouldn’t it be far more consistent and effective?

Perhaps someone read the word programming and thought: brain = computer.

Sure. A pre–Windows 95 computer, complete with blue screens.

You use what works. Not what promises miracles.

And finally, the part that actually annoyed me:

The idea that Ericksonian hypnosis is only for intelligent minds.

Milton Erickson was a genius. Many of us study him for years and still fall short.

What he worked with was not intelligence. It was:

metaphor

ambiguity

flexible language

Ericksonian work is for creative minds, for rich inner worlds, for symbolic thinkers.

People like those who read and write on Tumblr.

The world is full of idiots who are highly hypnotizable. And full of creative people who make excellent subjects.

Intelligence is not the requirement.

Structure is.

Oh yeah, it’s absolutely a crock of shit.

Listen, any time you go under general anesthesia is risky, and the longer you’re under, the more risks there are. So a surgery that takes 6-8 hours, no matter what it is, is going to be potentially dangerous.

But it’s important to know that a risk being present doesn’t make it LIKELY. You take at least as much risk driving to the grocery store as you would having phalloplasty—and if lightning strikes and something life-threatening DOES happen to you on the road, you’re not going to have an entire team of highly-trained, well-equipped medical personnel literally standing over you seeing to your wellbeing when it happens.

If you need a guide for making a choice (read below):

The Digital Media Myth We Need to Stop Believing

Digital media culture today relies on screenshots as the definitive currency of truth. People treat these images like absolute evidence during celebrity scandals, political squabbles, and viral gossip sessions. Someone posts a clipped image of a chat log or a deleted tweet, and the internet essentially declares the mystery solved instantly. Investigation fades away. Questions become obsolete. Everything reduces down to collective vibes and performative outrage. Truth be told, this constant reflex to treat pixels as judicial authority remains a massive oversight in our modern habits. That is the reality. Accepting screenshots as objective proof is practically reaching for a mirage while dying of thirst. Users habitually forget to interrogate the actual origins of what they are viewing. What if the image is fake, meticulously cropped, or missing the critical context required to understand the exchange? Navigating the chaotic environment of digital misinformation requires recognizing that these files change hands faster than anyone can verify the underlying reality. They represent a dangerous shortcut in an age that demands patience. Digital platforms like Instagram, WhatsApp, or X operate on specific visual dialects that our brains recognize instantly. When we see a familiar bubble design, a verified checkmark, or a standard timestamp, our cognitive biases latch onto that as an objective truth. It feels genuine because it mimics our daily digital lives.

Fakery succeeds because of this psychological familiarity. Reactivity dominates the online landscape. Most users consume information in a blur and react with raw emotion before stopping to analyze the data. Content that triggers anger, shock, or empathy travels at light speed across social feeds compared to dry, nuanced explanations. Algorithmic rewards for engagement prioritize these high-tension clips, turning dramatic, misleading screenshots into the primary drivers of internet traffic. It just works. Adding modern AI tools to this mix creates a landscape where fabrication becomes effortless. Sophisticated software is no longer a requirement; simple web generators turn anyone into a visual forger within moments. Changing a username, altering a timestamp, or inventing an entire conversation occurs in under sixty seconds. Beyond outright fabrication, cropping remains a pervasive weapon against the truth. A screenshot captures a single, frozen second while intentionally omitting the broader, messy picture. Imagine seeing a person yelling back at someone in a text thread. Without visibility into the preceding insults, that person appears unhinged and aggressive. Proper context dictates the true nature of communication, yet these images thrive on what experts call context collapse.

Reputations crumble daily because of these singular, static images used to fuel hate campaigns. By the time a full record emerges to tell the accurate story, the internet has already moved to the next controversy, leaving the original victim labelled by a falsehood. People latch onto the initial scandalous claim far longer than they attend to any subsequent, boring correction. Recurring issues like sharing five-year-old tweets as if they happened ten minutes ago further muddle our collective history. Parody accounts frequently trick gullible users who never bother to click the actual profile. Journalism suffers as well, with media outlets often reprinting screenshots as content instead of chasing down real sources and verified documentation. This creates a feedback loop where misinformation gains a veneer of legitimacy through repeated exposure. Artificial intelligence only deepens the divide by making images easy to fabricate, meaning our old intuition about visual evidence essentially holds no weight anymore. Deepfakes and synthetic conversations now blur reality to the point of exhaustion. Many individuals choose to trust a redacted, grainy screenshot from some random troll account over a multi-sourced, professional report. It turns out that velocity and sensationalism now dictate our standards for belief more than anything else.

Scrutinizing evidence is a dying art. Not every screenshot is a lie, but not every one is a holy document, either. Activists and researchers often save these files as necessary records before companies delete active data. They provide windows into genuine harassment or evidence of institutional corruption. The flaw is not the feature, but the blind trust individuals place in a single screen capture. Treating a file as a starting point for inquiry is mandatory, but treating it as a closing statement is reckless. Becoming a smarter consumer requires specific habits. Check the source first. If a faceless account with zero credibility pushes the image, skepticism remains the only logical path. Verify the account. Check the formatting. Scan for inconsistencies in the fonts or the spacing that might suggest a digital edit. Does the information appear on established news wires or other independent platforms that avoid the trap of unverified viral drama? If not, remain cautious. Slowing down represents the most critical act a user can perform. Emotional contagion on social media is a design feature, not an accident, and resisting that pull requires discipline.