Every document I read at work has been AI generated at some capacity. It's becoming so annoying that I have to use AI to create a summary.
I don't know if anyone agrees, but AI generated documents are so verbose and explain things in an overcomplicated way that makes me exhausted after reading. Sometimes I don't even understand what AI is trying to explain
> in an overcomplicated way that makes me exhausted after reading
This. I can't even listen to youtube videos any more because of their chat-gpt generated scripts. LLMs will readily generate several pages of superfluous prose, rehashing the same idea with punchy witticisms for no reason. But it uses such compelling connecting words that you might be tricked into believing there will be a payoff long past the point of diminishing returns.
I have been enjoying reading literary classics for their poetic concision. Sometimes, a bit of context can allow a simple sentence to form a picture in your mind's eye that would take a few minutes to explain syllogistically. I wish to communicate more like that.
I know you guys are joking, but this is like peeing in the sauna or shitting upwind. These serious writing samples will just enfreshen the training data.
No kidding. I got a feature handed down (with links to 4 other markdown docs) for what was discussed in person as "api to record audit events". The result is <goes to AI for word count then gives up>
Why do you think that? We’ve had this exact problem since the GPT3 moment and it has not improved at all. Not even a little, unlike agentic coding. Why do we expect a horizontal line to suddenly go exponential?
Also, it seems that writing encodes far more information about personality and mental state of the writer than it seems at first glance. A key problem is that machine-generated writing does not have this, it is obvious that it doesn’t have this, and there is no path towards fixing it.
Yea, this is definitely a place where the positive future outlooks have to state why.
The cost for text generation will trend toward zero.
The amount of models that will be able to generate text will keep expanding making what text is generated or not will become more difficult.
The way actual humans write will change because of influence by AI mushing up what is human and what is not human further increasing the impossibility of determining what is AI and what is not.
Freedom of speech laws in the US will shoot down any broad legal requirements to mark text as AI generated or not.
The consumer financial incentives to go to the 'better' and more expensive products are questionable at best. People consuming a lot of shitty shit because it's cheap is a pretty standard human behavior. It doesn't mean good products won't exist. It means good products will have to do a shit load of work to stand above and beyond the slop, and with systems like the US, these better products are apt to be bought by VC to exploit as much money out of the product by turning it into slop and getting rid of the human assets behind it.
Also..
You only notice 'slop', not necessarily things that are written by AI (or lazy humans).
Those people will stop really understanding anything because they defer all their writing (and so thinking) to machines. They will get stupider and stupider.
It's happening everywhere. We have people using AI to generate documents, then using the same AI to read/summarize them. What is the point? The writing itself, especially with recent Claude models, is often incredibly awkward.
It's the wet dream of any CEO. Subscription model for companies to pay to do tasks and subscription model for companies to understand the output of the subscription model. Amazing, they should do another Silicon valley season
The logical answer is to send what you would have put into AI.
However, terse lists of bullets without any padding or phatic expressions are considered rude. Perhaps previously a long message indicated you'd put time and thought into it.
Maybe AI will invert the social signalling, and make conciseness polite. As it should be.
What is the point is a question to a whole lot of human practices outside of AI, well, that I should say that AI is pointing out.
Huge numbers of us have mostly useless jobs. That's not to say they are completely useless, but maybe like at most 50% of the job is useful. The rest just gets filled with busy work and worthless text generation and TPS reports.
Don't romanticize the internet. The modern internet has torn society apart at the seams. It's made the world less tolerable and made civic discourse impossible.
If AI forces people to re-weight their priorities (i.e. be less on the internet), perhaps that's a good thing?
Yea, that internet was great, but that internet wasn't real. Or to put it in a better way, that internet wasn't in equilibrium with reality.
While social media may have ruined the internet, think of that as more of hitting the ground after falling off a cliff. It was the constant connection that really started changing everything. Instead of getting on the internet and disconnecting from the internet, we were always on. The algorithm that filled this new environment was inevitable.
Simply put there is no route back to those early days.
That's what happened to me and I'm happier than ever.
For the past five years I made an effort to seek out and finish long posts or videos that seem interesting at first in the name of exercising my attention span.
I'm now quick to bail on anything when I don't know the author, and bail on authors if their quality starts slipping. When the daily fire hose started running out, I realized I was mainly looking for distraction and spent more time being healthily bored.
The problem is AI and the internet's leakage into everything else. Marketing copy, prose in public spaces, even books... I really hope that society as a whole accepts that human-written text is more valuable than AI-written.
> Don't romanticize the internet. The modern internet has torn society apart at the seams. It's made the world less tolerable and made civic discourse impossible.
Pretty much the same thing the printing press did. Breakthrough communication technologies force underlying societal problems to the surface. It will end up being a good thing long term as long as we survive it.
I have had this thought quite a bit, that the silver lining of the slop-ification of the internet could lead people to spend less time online, which in my mind would be a positive development.
But unfortunately I just don't think that will happen. For some people it might, but I think the mass of people will accept the slop as it's given to them. Nothing in the history of communication technologies makes me think otherwise.
I don't know where this nostalgia is coming from. "The Internet" has never been this super perfect place on a pedestal some people make it sound like. The Internet is what you make of it. Not more, not less.
I was there in the early days (90's, early 2000's)
It was categorically, undeniably better--that's where the nostalgia is coming from.
There are lots of reasons why this happened (and those are interesting questions.) But simply denying that it was better doesn't seem particularly reasonable to me.
I miss the early days. I made my first "telnet" connection in either 1990 or 1991. I remember exploring gopher, FTP, early web sites, the transition from BBSes to early ISPs. It was all so exciting and new.
The Internet of today is barely recognizable. The very thing that made it sustainable (commercialization) is also what led to its demise.
There are still sites (like HN!) that give me a taste of the old world. But they are few and far between.
Constraints tend to make you enjoy something more. I agree that a lot of the things that annoy me today weren't present back then. But a lot of other things that I take for granted these days also weren't present back then.
Also, it is not a monolith. There are parts of the internet (arguably like HN) that are off the charts when it comes signal-to-noise ratio, for me personally.
If people are being honest, the nostalgia comes from a time where people only had to interact with other high-income, high-education, broadly-cosmopolitan technology enthusiasts similar to themselves. The moment the internet become used by everyone, including the riff-raff, it suddenly become "not fun".
ZKPs (as well as C2PA and similar user attestation systems) are quite literally a DRM scheme for the web, fraught with all the usual control and accessibility issues while still being relatively easy to circumvent by a sufficiently motivated and funded actor. Given the huge number of shady companies that run "phone farms" for purposes ranging from social media spam to Spotify royalty farming, I doubt they would even put a dent on this kind of automated abuse. They would however be very effective at locking every Linux and GrapheneOS user out of the internet.
Depends on what you define "internet" as, and what you expect it to be. So if it becomes a garbage, probably that's the equilibrium state it could get to, due to the forces around it. There are no isolated phenomena.
Slop is awful, but I've found networks like Mastodon and Bluesky where most users are actively antagonistic to still be enjoyable. Beyond that I've taken to leaning on RSS, private chats and automated filtering where it makes sense. Folks are setting up blogs again too. Most of the bigger platforms are flooded with slop, but their owners are encouraging that. Those spaces weren't healthy to begin with and they aren't now, but the reasons are different.
The never ending catastrophizing of every little thing is going to be the greatest cause of societal collapse. It's sucking the joy out of every aspect of life and it's usually some idiot hyperbolizing that most people don't grasp because most people don't even grasp sarcasm now because it's been so overused. It really wears on me and makes me think that Most people just want life to be horrible.
There are two conflicting messages here: AI is ruining the Internet, but also now the relationship with the Internet is more "peaceful".
A more peaceful relationship with the internet does sound good, and most of us probably waste too much time with social media in some form (including here).
My "optimistic" take on this is that AI will ruin the Internet and cause us all to spend more time with people in real life. And maybe that's a good thing.
That's assuming that AI Internet slop is such a turn-off that most of us will decide to disengage from it, which is a big assumption. But I did say it was optimistic.
What really makes my internet use worthwhile is interest-based tight knit communities that also interact in meat-space. Members are vetted based on personhood (REAL personhood) , not points/likes/bullshit.
Pay attention to what you consume. Find sources you trust, creators you enjoy, and communities that are real. Participate.
Not sure what else to say. Yeah, if you consume the "internet" like some firehose of information you spray at your face hoping for the best then, yeah, even today you're going to wind up gargling mostly shit.
The "internet" being a small group of well-minded people who largely shared a similar outlook on things was the anomaly, not what's going on now.
While I disagree with The Dark Forest hypothesis, the way AI and LLMs interact with the internet validate the theory in this specific instance. Individuals hosting their own slice of the internet have to deal with crawlers DDoSing them into oblivion and other islands retreat into privacy to avoid them entirely.
I'm beginning to wonder if one resolution to the Fermi Paradox is an Ouroboros loop of personalized content tailored to be exactly what one craves. No need for hunter killer robots, or global pandemic, just a gradual painless decline into virtual solipsism ought to get the job done. Brawndo for the mind.
Directionally I think social media is going to be regarded like tetraethyl lead and asbestos. (So is a lot of the ways we use plastic, though maybe not actual e.g. polyethlyene itself)
In the short term, I think AI may do something like this, but in the longer term, I fear that its capacity for propaganda and surveillance will produce something even worse than social media.
This is another one of those "duh" posts that many people seem to be slowly waking up to while the python of AI wraps around the body of the internet and squeezes the life out of it in preparation to shove the whole thing down its gaping maw. Still, everyone will go on about how great vibe coding is, and LLM models and sing "Nearer my God To Thee".
It's not just the internet, it's everything.
Every document I read at work has been AI generated at some capacity. It's becoming so annoying that I have to use AI to create a summary.
I don't know if anyone agrees, but AI generated documents are so verbose and explain things in an overcomplicated way that makes me exhausted after reading. Sometimes I don't even understand what AI is trying to explain
> in an overcomplicated way that makes me exhausted after reading
This. I can't even listen to youtube videos any more because of their chat-gpt generated scripts. LLMs will readily generate several pages of superfluous prose, rehashing the same idea with punchy witticisms for no reason. But it uses such compelling connecting words that you might be tricked into believing there will be a payoff long past the point of diminishing returns.
I have been enjoying reading literary classics for their poetic concision. Sometimes, a bit of context can allow a simple sentence to form a picture in your mind's eye that would take a few minutes to explain syllogistically. I wish to communicate more like that.
For too long we have used length of document as a measure of quality.
Your comment’s load-bearing property, asserts on the geometry that states it.
You've hit on something important, and it's real, not cosmetic.
I know you guys are joking, but this is like peeing in the sauna or shitting upwind. These serious writing samples will just enfreshen the training data.
This was really well written, and you hit on a serious point that most miss
Claude's butthole logo, meet Idiocracy
[dead]
A little sick in the throat is normal.
[dead]
No kidding. I got a feature handed down (with links to 4 other markdown docs) for what was discussed in person as "api to record audit events". The result is <goes to AI for word count then gives up>
Most likely outcome seems to be that this period will be over very quickly, and we will have better writing than ever before, forever.
We should probably put some more thought toward that part and what it will mean.
Why do you think that? We’ve had this exact problem since the GPT3 moment and it has not improved at all. Not even a little, unlike agentic coding. Why do we expect a horizontal line to suddenly go exponential?
Also, it seems that writing encodes far more information about personality and mental state of the writer than it seems at first glance. A key problem is that machine-generated writing does not have this, it is obvious that it doesn’t have this, and there is no path towards fixing it.
What exact problem? Verbosity? Quality?
The quality is eroded by the verbosity
[dead]
Why is that the most likely outcome?
The history of automation is filled with cases where the result has been lower quality goods at lower prices.
Yea, this is definitely a place where the positive future outlooks have to state why.
The cost for text generation will trend toward zero.
The amount of models that will be able to generate text will keep expanding making what text is generated or not will become more difficult.
The way actual humans write will change because of influence by AI mushing up what is human and what is not human further increasing the impossibility of determining what is AI and what is not.
Freedom of speech laws in the US will shoot down any broad legal requirements to mark text as AI generated or not.
The consumer financial incentives to go to the 'better' and more expensive products are questionable at best. People consuming a lot of shitty shit because it's cheap is a pretty standard human behavior. It doesn't mean good products won't exist. It means good products will have to do a shit load of work to stand above and beyond the slop, and with systems like the US, these better products are apt to be bought by VC to exploit as much money out of the product by turning it into slop and getting rid of the human assets behind it.
Also..
You only notice 'slop', not necessarily things that are written by AI (or lazy humans).
> Why is that the most likely outcome?
Listen, very smart people are very enthusiastic about other very smart people using AI to help AI learn to fix AI. Pretty much open-and-shut, IMO.
Those people will stop really understanding anything because they defer all their writing (and so thinking) to machines. They will get stupider and stupider.
We're coming up on 4 years into the AI-slop season of the internet. Just one more year and then we all go extinct?
It's happening everywhere. We have people using AI to generate documents, then using the same AI to read/summarize them. What is the point? The writing itself, especially with recent Claude models, is often incredibly awkward.
It's the wet dream of any CEO. Subscription model for companies to pay to do tasks and subscription model for companies to understand the output of the subscription model. Amazing, they should do another Silicon valley season
The logical answer is to send what you would have put into AI.
However, terse lists of bullets without any padding or phatic expressions are considered rude. Perhaps previously a long message indicated you'd put time and thought into it.
Maybe AI will invert the social signalling, and make conciseness polite. As it should be.
>What is the point?
What is the point is a question to a whole lot of human practices outside of AI, well, that I should say that AI is pointing out.
Huge numbers of us have mostly useless jobs. That's not to say they are completely useless, but maybe like at most 50% of the job is useful. The rest just gets filled with busy work and worthless text generation and TPS reports.
I like to call this phenomenon "reverse compression", or occasionally "depression". It's even lossy and subject to generational loss.
Hah. "Depression" fits well for a few reasons.
This can be improved. As AI grows it will be better at explaining. For sure people working on frontier models have identified this as a problem.
They didn't or they can't fix. Claude opus 5 writing is much worse than 4 years older models like gpt4
[dead]
It's still very early days. I have no doubt that etiquette and norms will be established that alleviate these issues.
Don't romanticize the internet. The modern internet has torn society apart at the seams. It's made the world less tolerable and made civic discourse impossible.
If AI forces people to re-weight their priorities (i.e. be less on the internet), perhaps that's a good thing?
You're describing the 2010s era of algorithmic social media, I think OP is referring to the earlier internet
Yea, that internet was great, but that internet wasn't real. Or to put it in a better way, that internet wasn't in equilibrium with reality.
While social media may have ruined the internet, think of that as more of hitting the ground after falling off a cliff. It was the constant connection that really started changing everything. Instead of getting on the internet and disconnecting from the internet, we were always on. The algorithm that filled this new environment was inevitable.
Simply put there is no route back to those early days.
>Simply put there is no route back to those early days.
Sure there is. Make another internet and keep the normies and capitalists away from it: https://news.ycombinator.com/item?id=49541705
That's what happened to me and I'm happier than ever.
For the past five years I made an effort to seek out and finish long posts or videos that seem interesting at first in the name of exercising my attention span.
I'm now quick to bail on anything when I don't know the author, and bail on authors if their quality starts slipping. When the daily fire hose started running out, I realized I was mainly looking for distraction and spent more time being healthily bored.
>be less on the internet
you might as well expect people to drink less water.
The problem is AI and the internet's leakage into everything else. Marketing copy, prose in public spaces, even books... I really hope that society as a whole accepts that human-written text is more valuable than AI-written.
> Don't romanticize the internet. The modern internet has torn society apart at the seams. It's made the world less tolerable and made civic discourse impossible.
Pretty much the same thing the printing press did. Breakthrough communication technologies force underlying societal problems to the surface. It will end up being a good thing long term as long as we survive it.
I have had this thought quite a bit, that the silver lining of the slop-ification of the internet could lead people to spend less time online, which in my mind would be a positive development.
But unfortunately I just don't think that will happen. For some people it might, but I think the mass of people will accept the slop as it's given to them. Nothing in the history of communication technologies makes me think otherwise.
Perhaps I'm too cynical.
be careful for what you wish for. if you take away our drug/escape/safety. The people will switch from #revolting to 1800-1900s revolting.
The Internet began to break at the same day you gave it away to Jeff Zuckermusk.
NOW people start to complain. And it's quite transparent why. ;)
I don't know where this nostalgia is coming from. "The Internet" has never been this super perfect place on a pedestal some people make it sound like. The Internet is what you make of it. Not more, not less.
I was there in the early days (90's, early 2000's)
It was categorically, undeniably better--that's where the nostalgia is coming from.
There are lots of reasons why this happened (and those are interesting questions.) But simply denying that it was better doesn't seem particularly reasonable to me.
I miss the early days. I made my first "telnet" connection in either 1990 or 1991. I remember exploring gopher, FTP, early web sites, the transition from BBSes to early ISPs. It was all so exciting and new.
The Internet of today is barely recognizable. The very thing that made it sustainable (commercialization) is also what led to its demise.
There are still sites (like HN!) that give me a taste of the old world. But they are few and far between.
It was better because it was, well, less.
Constraints tend to make you enjoy something more. I agree that a lot of the things that annoy me today weren't present back then. But a lot of other things that I take for granted these days also weren't present back then.
Was it better? No, I don't think so.
Agreed.
Also, it is not a monolith. There are parts of the internet (arguably like HN) that are off the charts when it comes signal-to-noise ratio, for me personally.
If people are being honest, the nostalgia comes from a time where people only had to interact with other high-income, high-education, broadly-cosmopolitan technology enthusiasts similar to themselves. The moment the internet become used by everyone, including the riff-raff, it suddenly become "not fun".
I like duckduckgo, I find the search assist to be useful, and afaict it doesn't harvest your data.
Why can't the post-ai internet include personal agents that work for you or at least don't make you the product?
The same people who would petition against ZKP Internet IDs complaining about the unregulated internet.
You can't have your cake and eat it too...
ZKPs (as well as C2PA and similar user attestation systems) are quite literally a DRM scheme for the web, fraught with all the usual control and accessibility issues while still being relatively easy to circumvent by a sufficiently motivated and funded actor. Given the huge number of shady companies that run "phone farms" for purposes ranging from social media spam to Spotify royalty farming, I doubt they would even put a dent on this kind of automated abuse. They would however be very effective at locking every Linux and GrapheneOS user out of the internet.
Depends on what you define "internet" as, and what you expect it to be. So if it becomes a garbage, probably that's the equilibrium state it could get to, due to the forces around it. There are no isolated phenomena.
Slop is awful, but I've found networks like Mastodon and Bluesky where most users are actively antagonistic to still be enjoyable. Beyond that I've taken to leaning on RSS, private chats and automated filtering where it makes sense. Folks are setting up blogs again too. Most of the bigger platforms are flooded with slop, but their owners are encouraging that. Those spaces weren't healthy to begin with and they aren't now, but the reasons are different.
The never ending catastrophizing of every little thing is going to be the greatest cause of societal collapse. It's sucking the joy out of every aspect of life and it's usually some idiot hyperbolizing that most people don't grasp because most people don't even grasp sarcasm now because it's been so overused. It really wears on me and makes me think that Most people just want life to be horrible.
> It really wears on me and makes me think that Most people just want life to be horrible.
Yet you are the one that sounds miserable.
There are two conflicting messages here: AI is ruining the Internet, but also now the relationship with the Internet is more "peaceful".
A more peaceful relationship with the internet does sound good, and most of us probably waste too much time with social media in some form (including here).
My "optimistic" take on this is that AI will ruin the Internet and cause us all to spend more time with people in real life. And maybe that's a good thing.
That's assuming that AI Internet slop is such a turn-off that most of us will decide to disengage from it, which is a big assumption. But I did say it was optimistic.
I fear that slop will end up being hi quality ai videos that will hook the masses.
That is probably the more likely scenario.
If places cease to be interesting, people will stop visiting them.
That means either AI will become interesting, or people will figure out how to stop slop ruining their places.
Either of those options might bring their own problems, but the sloppification of the internet is not a stable equilibrium.
Is there no-AI internet? Like anti-nuclear shelters?
What really makes my internet use worthwhile is interest-based tight knit communities that also interact in meat-space. Members are vetted based on personhood (REAL personhood) , not points/likes/bullshit.
Pay attention to what you consume. Find sources you trust, creators you enjoy, and communities that are real. Participate.
Not sure what else to say. Yeah, if you consume the "internet" like some firehose of information you spray at your face hoping for the best then, yeah, even today you're going to wind up gargling mostly shit.
The "internet" being a small group of well-minded people who largely shared a similar outlook on things was the anomaly, not what's going on now.
There is Gemini (the real gemini) and gopher, so far there is a lot of interesting items. Some Links:
gemini://gem.sdf.org/
gemini://sdf.org/
gemini://gemi.dev/
gemini://gemi.dev/cgi-bin/wp.cgi/
gemini://cdg.thegonz.net/infrastructure/search/
gemini://zaibatsu.circumlunar.space/~solderpunk/cia-world-factbook/
gemini://geminiprotocol.net/
gopher://sdf.org/1/
gopher://magical.fish/1/
gopher://gopherpedia.com/1/
gopher://gopher.floodgap.com/1/overbite
The real gemini has always been https://www.gemini.com/, the crypto exchange founded by Mark Zuckerberg's twin nemeses, the Winkelvii
While I disagree with The Dark Forest hypothesis, the way AI and LLMs interact with the internet validate the theory in this specific instance. Individuals hosting their own slice of the internet have to deal with crawlers DDoSing them into oblivion and other islands retreat into privacy to avoid them entirely.
I'm beginning to wonder if one resolution to the Fermi Paradox is an Ouroboros loop of personalized content tailored to be exactly what one craves. No need for hunter killer robots, or global pandemic, just a gradual painless decline into virtual solipsism ought to get the job done. Brawndo for the mind.
I would say what he means is the web. LLMs can be used as another interface to the internet. Additionally the web gets polluted by slop.
If AI slop completely kills social media then that’ll only be a good thing. Bring it on down.
Directionally I think social media is going to be regarded like tetraethyl lead and asbestos. (So is a lot of the ways we use plastic, though maybe not actual e.g. polyethlyene itself)
In the short term, I think AI may do something like this, but in the longer term, I fear that its capacity for propaganda and surveillance will produce something even worse than social media.
There's so much spam and junk everywhere. It's all slop now.
This is another one of those "duh" posts that many people seem to be slowly waking up to while the python of AI wraps around the body of the internet and squeezes the life out of it in preparation to shove the whole thing down its gaping maw. Still, everyone will go on about how great vibe coding is, and LLM models and sing "Nearer my God To Thee".