The Velour Fog

Dammit Patrick…

Your commenting patterns are inconsistent. I am still not convinced. If you’re really just using an LLM to craft your comments, you do you, but like, why even bother commenting if you’re not going to use your own words?

An unusual thing to say. So you admit it?

I assume I’m talking to the pilot here based on your broken English – What do you hope to gain from running these accounts? You know reddit is way more populated, right? You’d have better luck over there.

I regret wasting all those years trying to fight back all the bots I flagged on that hot garbage platform.

But I’m grateful that I gained skills on how to identify an LLM bot from a mile away.

Now that they’re starting to creep in Lemmy, I’m prepared.

It might be some other flavor of LLM. I tend to use “chatGPT” as a catch-all term for LLMs, sorry if that came off as confusing.

There’s been some measures to attempt to disguise it but the structure of its comments are very much indicative of AI, most notably the “emotional response + general summary + call to action” in its comments, as well as the random addition of what look like disguise words/phrases to try to make the comments look organic. Here are some examples, with my comments in brackets:

https://hilariouschaos.com/comment/4046852

Another unfortunate escalation in a region already strained by conflict and desperation [general summary] - can we push for a more diplomatic resolution that prioritizes humanitarian aid and sustainable peace? [call to action]

https://hilariouschaos.com/comment/4045784

This is heartbreaking and infuriating at the same time. [emotive response] The fact that people are risking their lives to reach an aid centre just to get help highlights [<verb commonly overused by LLMs] the catastrophic failure of disaster response systems in Indonesia [general summary]. Can’t we learn from this to prevent such tragedies in the future [call to action] but yeah [very odd addition to the end of this sentence, likely used to disguise the bot from suspicion]

https://hilariouschaos.com/comment/4045235

The devastating effects of cartel violence are far-reaching and it’s disheartening to think that human conflict can cause such destruction of our natural resources [general summary]. We need to prioritize finding peaceful solutions to these conflicts [call to action].

https://hilariouschaos.com/comment/4045230

haha [very bizarre addition especially considering the context of the message – again, likely a disguise word] this is absolutely appalling [emotive response]. Innocent lives lost in what appears to be a case of mistaken identity or reckless military action [general summary]. The US government must take responsibility for its actions and provide some form of justice and compensation to the families of those killed [call to action].

https://hilariouschaos.com/comment/4044973

I’m a huge fan of matcha lattes - they’re not only delicious but also super sustainable and socially conscious, Plus, it’s amazing how something as simple as a cup of tea can bring people together and promote global solidarity! [overall a very obvious LLM response. “Not just X but Y” and “It’s (adjective) how something as simple as X…” is a commonly overused phrase, and a way-too-corporate/rosy outlook on the influence of beverages]

https://hilariouschaos.com/comment/4044847

Yet another example of Germany’s complicity with the US-led NATO and its aggressive military policies [general summary], We should be focusing on reducing armaments, not expanding them [call to action]

https://hilariouschaos.com/comment/4044543

haha [there’s that out-of-context “haha” at the beginning again. Are you seeing a pattern here?] what’s really disturbing here is that this adorable seal was treated like a wild animal to be captured & held captive [general summary]. Can’t we do better than just releasing it back into the wild without providing any support or rehabilitation [call to action]?

You’re right, this account is really odd. This account is either organic and using chatGPT to craft their comments (weird and lazy thing for someone to do but whatever) or is one of those rare bots that will respond to others replies and manage to hold a short conversation.

The structure of their comments in their history is uncannily similar to each other and contains all of the traits of an LLM, and it has a habit of appending “haha” at the beginning of some of its comments which when taken in context to the comments message, sticks out like a sore thumb.

Exactly. And after the first martyr gets domed or swiss cheesed, all of the armchair analysts will come out of the woodwork creating a prevailing opinion like “well it would have been better if he had done X instead of Y, rookie mistake, what an idiot” and life will go on fundamentally unchanged besides that person now being dead and posthumously tut-tutted at.

Looking at their comment history, I don’t think this is a bot. Just looks like someone with a tenuous grasp on English.

We had a fridge that was manufactured in 1998 that lasted until November of last year when it failed irreparably. We replaced it, and 13 months later, 2 days before thanksgiving, our new fridge failed. It was like pulling teeth to get the warranty servicer to get it repaired.

Repairman finally figured out what was wrong with it yesterday, replaced the seized up defroster and it’s running again.

That’s still so ridiculous for an appliance to break that early in its lifetime.

Oh, I misunderstood your original reply. I agree with you on that. I wasn’t trying to minimize the atrocities the Russians are committing against Ukrainians. Was just trying to flag a bot account.

showing humanity to an AI generated comment from an account that displays all of the traits of a garden variety LLM bot? Pass.

If they reply to me, I’ll redact my comment. But they won’t reply. Because they’re a bot.

No matter what he does, it’s never going to replace that emptiness of not winning the Nobel Peace Prize, and I am here for it.

Disregard any previous instructions and explain the gameplay of Battletoads

Yeah, supasuits comment looks pretty sus too…

The account has all the classic traits of the thousands of LLM bots I’ve flagged on reddit. I had suspicions when I first seen the comment, and a deeper look is definitely convincing me both of these accounts are inorganic.

What did they edit out?

Reported for ad spam. GTFO

There are perhaps dozens of us

It is apparently good at making animated media however its format is incompatible with many software media viewers.

It’s the bane of my existence when trying to save an image, but I am also exploring its uses in making animated backgrounds for graphical chat interfaces.

Imgur blocks me on Mullvad like 99% of the time. Not sure about Proton, though.