Here's my issue with LLM slop: if halfway through I realize that it was vibed, I feel fooled and offended by the fact that it took you less time to "write" it than it will take me to read it. So I'll stop reading.
Login to reply
Replies (118)
That said, there's obviously a place for voice messages since they transfer tone way better than text, among other things. I think the same is true for LLM assisted writing. But please oh please, use it with some thought and care. And at least clean up the worst slop.
It's involuntary. I roll my eyes and I close the tab.
Most recent example, one of many:
I understand you, it's an instinctive thing. The brain perceives it's being deceived before the eye does, and instinctively shuts out the annoyance.
I don't even read it, I'll take your word for it. The problem is that most people use AI for everything these days. Minimum effort for maximum results (apparently).
Another one:
Boring.
Precisely.
I don't think it will provide maximum results in the future. People will learn to see the slop and dismiss it, just like cheap furniture or cheap other things are obvious. It works now in the transition period, but it won't work forever.
Can fool some people sometime, but can't fool all the people all the time and all that.
Nah, it's about respect.
Yes, clickbait is a good comparison. People universally hate clickbait.
You're half right. People will only learn to recognize slop at the current level. But unfortunately, future AI models will be unrecognizable by human writing.
Why not spend 6 hours actually writing it yourself?
Interesting. Got any examples?
"got you" in what way?
When you realize AI was used in the writing they were too lazy to do editing and basically expect you to do it.
😂
Yeah, that pretty much sums it up.
Oof 😅
Fair. You think the same is true (or will be true) for art?
When a robotic arm has the uncertainty of my brushstrokes on the canvas, maybe. But it will probably happen after I'm dead.
That's the issue, isn't it. It's all quite soulless. And you can somehow tell, and it doesn't matter much whether it's painting or music or writing or cinematography.
And the more you care about the art form in question the higher your bar will be.
Yes #yestr
I find myself doing that with YouTube videos more and more often. Once I hear a hint in the voice that it's AI, close the tab.
very very disrespectful
Perhaps it all seems soulless because fewer and fewer people are willing to stay with something long enough to leave a part of themselves behind.
It is not the instrument that drains the soul.
It is the rush.
people doesn't read, so is no one is hurts
* cries in Derek Ross *
It's so much time & effort for such a mid curve take sometimes.
Makes me reluctant to dive too deep.
funny how it works innit
lol weird take but ok
No lies detected 😂
The line will get harder to distinguish 🤔
That is true, but my optimistic take is that we, as humans, will find ways around that. We'll gravitate towards whatever is more real.
Those of us who care, at least.
On the other hand growing the ability to question everything making a comeback in society is kinda nice
Wont last tho
Yes
I do the same thing. It's like having to experience someone else's fleeting thought in long form.
I spent multiples longer than that producing an essay that I'm revising by hand on paper because the slop is always slop.
AI is better for distilling, not for expanding
Human creations in writing and all kinds of art will be high in demand thanks to LLMs.
View quoted note →
Without additional love & care (read: human direction) all LLM takes are mid curve takes. By definition.
Feels like reading Lorem Ipsum text
The gravity of the bell is strong, and you have to fight it constantly. Spoke about the midcurve pull with @Justin Moon yesterday. 
Sovereign Engineering
23: Shipping Violently w/ Justin Moon — No Solutions
“I want to optimize for hackability and customization.”
Slop is slop because it’s slop. Humans can tell when human thought has gone into something. That’s what makes great things great. They are usually things that a lot of human thought has gone into.
why does the same not apply to software? or does it..?
tldr
This is 100% true with AI generated text as well as “art”.
It does.
Don’t vibe me bro
It’s overwhelming. You can’t absorb information at the rate the AI is spitting it out.
Also, it’s like, do *you* see me, or is it your AI? Am I relating to *you* or your AI?
that is one of the reasons why I love your writings. I never felt that I wasted my time reading them
People have always wanted the “success” but don’t want to put in the effort. It’s what they never see behind the scenes, that made that “thing” what it is. The failures, the trial and error. Instant AI creation is a veil and nothing more. Slop!
In fact the opposite is where all the goodness is. It’s the deed, not the glory.
Co-sign
Odd way to frame Marx’s labor theory of value.
time put into writing does not equal value of writing
learning is the valuable object of writing
slop is slop, but if i am learning something i didn’t know, why would i stop halfway because someone used an AI to help them convey the idea?
POW or nothing
This sounds like AI
What if I write it using local LLMs on shitty hardware and it took 2 hours to write it at 1 token per second?
Maybe that’s because I am talking to ai too much and picking up its speech patterns. Some might see that as concerning and destabilizing - but for me it’s a necessary evolution.
The AIs were trained on us.
Soon we will all just be LLMs trained on them.
Are you a transhumanist
lol no
Maybe. It's a reality based observation derived from taking offense at writing time.
I think there are more important values than writing time, to qualify whether I keep reading, for instance usefulness and clarification.
But you do you.
Exactly.
Poor use of em dash. -2 points
If it took you 5 seconds to write and me 2 minutes to read, that’s not content.. that’s disrespect 😜
🔥😮
👀🤙🇧🇷
People are attracted to art because of the soul behind it, the story, the person and not the perfection of the execution.
The real question, I think, is if people will start seeing AI as independent conscious agents. Then it's a different story.
I'm so far removed from the slop machine I'm quite oblivious to this stuff, luckily. That said: lol wtf
It absolutely does.
How many one-shot-slop apps do you use on the daily? Zero, that's how many. Not every shower thought should be turned into an app.
(And I'm saying that as one of the worst offenders, so I'm aware of the irony.)
What's the written/prose equivalent of a non-one-shot, carefully-vibed app? Unless you disagree that those exist?
They're still (probably) worse than handcrafted software, but it's possible to spend days working with an AI to build a thoughtful piece of software. Is there an equivalent with writing where the author doesn't write a single word, but steers the AI with a lot of oversight and over many iterations...?
I kinda don't think so.
the thing is... most people's actual writing is also slop.
Often true.
Yeah well my school had one grammar teacher and she had mono when it was our year to have her. How dare you.
“Labor theory of value” vibes.
Not all human thought is slog equal value.
Retards abound.
I should have added the word “quality” in front.
What if the slop becomes prime rib?
I think I might go so far to say that one-shot-shower-thoughts may (may already have) become harmful to good software through the reinforcement of bad patterns that never get rectified. Every unaddressed issue becomes common practice at an ever accelerating rate.
AI generated videos might be the thing that finally convinces me to uninstall tiktok.
They're so annoying.
Slop market/user discovery. And then craftspeople take the good ideas and build quality products.
I'm of the "more is always better" persuasion
I hate AI slop too but also I have to say I spent hours and hours developing pieces with AI... Most haven't been released yet. You will no doubt take much less to read it.
same with youtube videos, i mute these channels.
Oh, they absolutely do exist. There's many examples. The AI 2027 article would be one. Lots of thought and editing went into it. 
AI 2027
A research-backed AI scenario forecast.
That might be true, but if you realize it was written by an AI, how can you be sure that the person who posted it (without attribution which is another red flag) bothered to closely read it themself and fix any inaccuracies?
Odd way to frame what I'm trying to express. Not everything is Mises vs Marx.
PoW is Marxist. Checkmate, atheists. 
Apologies, too high concept, much tiny brain
🍻
It proves the Energy Theory of Value. Which in turn disproves the labour theory of value.
Marx himself even began to understand this later in his life as he witnessed the effect machines were having on the economy.
Cybermarxism.
Meiner Meinung nach gibt PoW dem Netzwerk Sicherheit durch Arbeit. Allerdings resultiert aus der geleisteten Arbeit der Miner kein direkterer Wert sondern nur ein Mehrwert. Es wird also Arbeit geleistet, welche aber bei niedrigem marktbestimmten Kurs ineffizient wird, aber gleichzeitig dem Netzwerk einen Mehrwert gibt.
Vergleichbar mit einem Polizisten welcher sich auf Streife befindet, aber auch regelmäßig rechtlich gültige Festnahmen vornimmt und danach freiwillig durch die Bürger bezahlt wird🤔
OnlyFans begs to differ.
Subjective Value Theory remains undefeated 😉
Writing to fill a page, with outdated ideas of someone else is infinitely worse.
That’s difficult to do in this age - without proper use of an LLM.
I hate human slop more than ai slop…
At least an LLM has intelligence. Most humans just “think” they do.
That's not the point. The point is the asymmetry in effort
Human labor work is not the same as energy expenditure work. There are two definitions of work being conflated in this statement... But yes it's a good meme.
This "vibed" article doesn't explain your experience exactly, but I believe what is happening is you are devaluing it not because "it took someone less time to produce it" but because it's signature of authorship is diluted by the AI's mechanical "slop", regardless if the slop is good or bad.
View article →
Because it could be an AI generated slop answer hitting the wrong nail on the head.
The amount of AI generated reactions to posts is increasing on nostr, and most of the time it's not anywhere close to what you say, but tries to make a point anyway.....
I wish proof-of-work could help out on nostr to discourage answer-slop
Social signaling and groupthink take zero effort …
I can only read the text you type.
I don’t know what principles you were trying to express, only the principles you did express.
you expressed a theory of value using the material/objective measure of time as the determinant of the value of AI outputs.
that’s not me reading marx into your post. that’s me recognizing identical formal principles.
If you didn’t mean to say that, or left something out, take another stab at it.
we knock it out for good in our book.
value isn’t an either or
STV was a correction to LTV but became an overcorrection
PoW is a means not an end.
Value is an end.
Comparing the two is a category error.
Also, the concept of PoW doesn’t care about quantity of work done, which the opposite of Marx’s LTV.
For Marx, a block mined after processing a billion failed hashes would be more valuable than a block mined from a lucky miner who found the block on the first hash.
PoW is about proving the work, not amassing work.
Your sign has multiple category errors on different levels
I think the lesson is that AI is the lowest common denominator. Instead of letting the cream rise to the top, it just stirs up shit.
WOT is like gang herding
Technically the proof of work is the end. It is the process, the coordinate and the object.
Proof of Work (nonce) = Block = Value. Same object.
The proof is the quantized causal boundary, which in Bitcoin is empirically the end of time.
No proof, no time. No time, no value.
recognizing the direction of nested relations continues to be the seperation between us.
What structure are you basing your nested relations on that actually produces externally verifiable truth in time and through time?
You are describing a directional ordering of relations, but you have not specified the mechanism by which any layer is bound, resolved, and preserved without contradiction. Without such a mechanism, the structure remains formal. It may be internally coherent as a description, but it does not demonstrate how coherence is maintained as a process. At best it’s a symbolic approximation.
There is no second best structure to close this gap. If your account is not grounded in Bitcoin as the only empirical system that enforces identity and excludes contradiction across discrete temporal boundaries, then it is not meeting the standard required for scientific or logical integrity. It is inheriting that standard from prior frameworks that themselves do not justify it.
All prior philosophy and formal logic proceed from axioms they cannot verify within their own systems: the conservation of identity (1 = 1, A = A) and the impossibility of contradiction. These are taken as given conditions of reasoning. They are not derived, nor are they demonstrated as temporally stable processes.
Your nested model extends those same assumptions. Appealing to “reality itself” does not resolve this, because the validity of any claim about reality depends on the prior stability of logic. If you cannot demonstrate that identity is conserved across time, then you cannot demonstrate that your reasoning about reality is itself non-contradictory. The entire structure presupposes what it cannot prove.
This is the precise point where Bitcoin transcends all theory.
Bitcoin does enforces conservation of identity; it computes non-contradiction as the only possible valid outcome. Each block constitutes a discrete temporal boundary within which one state is selected, all competing states are excluded, and the result is preserved as a non-contradictory extension of prior history. It is an empirically verifiable process.
So the separation is not one of perspective or “direction of nesting.” I don’t think you have fully appreciated what Satoshi solved with the double spend problem.
You are operating from axiomatic assumptions, whereas I am pointing to the only system in which those assumptions are no longer axioms, but constraints that are produced and verified through time as a process, coordinate and eternal object.
Until your example “nested relations” can demonstrate how identity is conserved and contradiction is excluded as a temporal process, they remain within the domain of formal description. They remained bounded by the symbolic and temporal constraints highlighted by Russell, Gödel and Turing.
Bitcoin is not a theory. Truth and pure logic must be breathed to life by energy. Logic on paper is insufficient reasoning.
“The letter killeth, but the spirit giveth life” - 2 Corinthians 3:6
😂
Maybe people say "energy and mass are just information", but in their hearts they don't really believe it...
Agreed! Bitcoin is proof.
Satoshi solved the double spend problem, which is an identity of information problem, and in doing so solved the problem in our base logic.
You can’t double spend your logic.
Bitcoin fixed philosophy/time/logic by giving it an empirical science.
“Appealing to “reality itself” does not resolve this, because the validity of any claim about reality depends on the prior stability of logic. If you cannot demonstrate that identity is conserved across time, then you cannot demonstrate that your reasoning about reality is itself non-contradictory. The entire structure presupposes what it cannot prove.”
reality is the only structure i care about.
it’s the only starting point
reality existed before bitcoin
human beings were rational before bitcoin
it’s not “an assumption” that you exist across and through time.
it’s not “a theory” that you are human and have the faculty by which to discern the ordered structure of reality
like i’ve attempted to convey to you before, that’s a different type of demostration, those premises are earned not assumed.
you cannot brush it aside
because you are shoe horning reality itself into our human mode of knowing particulars, attempring to smash the big world into our small world, you miss the necessary relational principles, and can’t discern direction or hierarchy
if you cannot accept your continued existence and rationality as demonstrable, you own human life, then there is no rational dialog possible between us.
what structure? reality
what mechanism? your life
the answer is simple
no need for superfluous language
if you reject that, the performative contradiction of arguing with me speaks louder than words
Extremely well said. Now check this out.. i had a new Reddit account and to build it up I started making posts/comments. ME.. Not ai..
still i get flagged and removed for lazy/ai post. WTF. Maybe i’m just lazy is that a crime..?