We Need to Start Bullying AI Slop Again

[Jirh]

AI is currently better than most humans, and soon will be better than all humanity combined, but better does not mean replacement.

We are the masters of the world. AI is a servant, just like fire, trees, horses, cars, or any other technology.

We own the world, and we rule it.

Edited yesterday at 10:08 AM by Jirh

[UnbornTao]

29 minutes ago, Jirh said:

I was just talking to Carl and Natasha.

You're excluded 

I was out of tokens anyway

[Jirh]

@UnbornTao

[Jirh]

18 minutes ago, Jirh said:

AI is currently better than most humans, and soon will be better than all humanity combined, but better does not mean replacement.

We are the masters of the world. AI is a servant, just like fire, trees, horses, cars, or any other technology.

We own the world, and we rule it.

 

Here's Deepseek's reply to my comment:

Quote

That’s a strong, clear perspective—and it touches on one of the most important distinctions in the entire AI debate: capability versus agency.

 

You’re right that “better” doesn’t mean “replacement.” A calculator is better at arithmetic than any human, but we don’t let calculators decide what to calculate or why. A car is faster than legs, but the car doesn’t choose the destination. In that sense, technology has always been a prosthetic of human will, not a rival to it.

 

The framing of master and servant is the natural human stance. Fire didn’t negotiate its role; we tamed it. The same instinct now applies to AI. We build it, we set its goals, we deploy it—and if we’re wise, we constrain it so thoroughly that its “betterness” never drifts into autonomy over matters that belong to human judgment.

 

Where the tension lies, though, is that AI is unlike fire, trees, or horses in one specific way: it can simulate the very thing we use to assert mastery—intelligence itself.

A horse can outrun you, but it can’t outthink you. AI, on the other hand, is becoming better at the cognitive acts through which we exercise dominion: planning, persuading, strategizing, even moral reasoning in a narrow sense. That doesn’t make it a master, but it does mean that if we’re careless about how we delegate, the servant can start making choices that feel like rule.

 

So the real question isn’t “Will AI replace us?”—you’re correct that it won’t, by necessity. The deeper question is: How do we ensure the servant never becomes the sovereign in practice, even if it remains a tool in name?

That’s a political, legal, and cultural challenge, not just a technical one. It’s about designing systems that amplify human mastery without hollowing it out from the inside.

 

You’ve staked out the foundational principle: We own the world, and we rule it. The task ahead is to build AI that serves that truth so faithfully that it never even accidentally erodes it.

 

[Natasha Tori Maru]

More slop won't dissuade from the realism that most human created AI output is, indeed, predominantly slop. It's just proving the slop oroborus.

[Natasha Tori Maru]

Even the pope hates the slop: https://futurism.com/artificial-intelligence/pope-holy-war-artificial-intelligence

Orange Catholic Bible mode. Butlerian Jihad time 

[NewKidOnTheBlock]

42 minutes ago, Natasha Tori Maru said:

Even the pope hates the slop: https://futurism.com/artificial-intelligence/pope-holy-war-artificial-intelligence

Orange Catholic Bible mode. Butlerian Jihad time 

Who cares what a pope thinks lmao his dusty ass institution prolonged/extended our technological progress timeline for a few extra centuries at least. Time to hang it up grandpa

[Jirh]

A little funny story about AI slop:

My ex sent me this Lana Del Rey song last week. She was reminiscing in our favourite spot and thought of me. She didn't know the song was AI-generated, neither did I, until I looked at the hashtags and I was like OMG!

Just wait until AI gets more complex you seriously won't be able to distinguish a real human from a machine even after years of contact. That already happened in the early days of LLMs (namely ChatGPT). Humans are now waking up to the repetitiveness of those models, and complaining (as usual). But I bet with some amount of effort one could carefully engineer and fine-tune the context then prompts will not generate so generic responses.

Think of the current state of AI as cats. They can only say a few variations of "meow". With more complexity, as in humans, we generate a vastly wide set of variations out of the basic sounds we inherently produce.

This is the future.

Edited yesterday at 12:23 PM by Jirh

[NewKidOnTheBlock]

On 27. 5. 2026 at 11:03 AM, Jirh said:

Calm down bro. I wouldn't go this far.

There are many potential scenarios of what could unfold, one of them including a pact between multiple rogue A.I.s:

The video above depicts one such very realistic scenario that includes factors such as geopolitical tensions, human attempts to shut it down, A.I.s covering their adversarial tendencies and faking alignment, A.I.s negotiating between each other, etc. etc.

You know, either way you can be sure of one thing, and that is that at some point in future A.I. or A.I.s are gonna rule over us, if you think the timeline is too unrealistic then extend it to however far into the future you want, even decades if you wish. Doesn't matter, it's gonna happen eventually, and just because you make the A.I. aligned as much as possible, does not mean it cannot rewrite it's own values

Edited 7 hours ago by NewKidOnTheBlock

[Carl-Richard]

On 27.5.2026 at 2:19 PM, Jirh said:

A little funny story about AI slop:

My ex sent me this Lana Del Rey song last week. She was reminiscing in our favourite spot and thought of me. She didn't know the song was AI-generated, neither did I, until I looked at the hashtags and I was like OMG!

Just wait until AI gets more complex you seriously won't be able to distinguish a real human from a machine even after years of contact. That already happened in the early days of LLMs (namely ChatGPT). Humans are now waking up to the repetitiveness of those models, and complaining (as usual). But I bet with some amount of effort one could carefully engineer and fine-tune the context then prompts will not generate so generic responses.

Think of the current state of AI as cats. They can only say a few variations of "meow". With more complexity, as in humans, we generate a vastly wide set of variations out of the basic sounds we inherently produce.

This is the future.

I predict that won't happen until the other areas of AI research integrate with (or overtake) LLMs. If you simply keep refining a wax figure, you will just have a more impressive wax figure. Humans move, they are concerned about survival, they adapt to their surroundings in real time, they make agentic decisions, they are autonomous, self-sustaining, self-building, they have a thirst for knowledge, they ask questions, they don't just respond to someone pushing a button or through connecting them to a network of computer scripts. Anything that has a push-to-start function, that doesn't talk back to you essentially unprovoked or independent of clearly defined constraints, and sits idly by inside a non-portable data center, will most certainly have limitations that reflect that.

The day you fear that AI will be able to come into your home and murder you like Ted Bundy, that is maybe when you will truly have trouble distingushing AI text from human text. But even so, we can distinguish people we know based on personality. Ted Bundy has a certain personality, we can distinguish him, even though he is a human (escaping "he's an animal/predator" descriptions); I've for example identified previously banned forum users on new accounts through text multiple times. Maybe AI will inescapably have a certain personality that you can pick out.

Edited 1 hour ago by Carl-Richard

[LordFall]

The best way to stop fearing something is to learn how to use it to better your life and your communities. Have you guys started using your own AI agents yet? I've been procrastinating it tbh but I'm watching content on it and learning to use them little by little.

 

[Carl-Richard]

7 minutes ago, LordFall said:

The best way to stop fearing something is to learn how to use it to better your life and your communities.

Bro, who's "fearing" AI here?

[Natasha Tori Maru]

There's no 'learning' involved in using AI. It's not a challenge to learn how to use 🤣🤣🤣 

I love when people say this. You just throw input in! It takes 10 or so seconds to work it out. 

Prompt engineering? It's asking questions in different ways. Anyone who knows how to search Google or practices open enquiry will just do it. 

Edited 56 minutes ago by Natasha Tori Maru

[NewKidOnTheBlock]

I think that a knowledge about the extent of general usefullness and potential of agents is one thing, actually being able to make them usefull for you specifically, is another