Nah, it’s saying that ad and AI-driven internet will prevail. People only use Google to find an answer and don’t dig deeper, and if they do, it’s often because the links are sponsored. People using GPT’s are even less likely to click a link. Currently no ads, but just wait.
“what should I do if I’m going through severe emotional distress? How to choose a good psychiatrist?”
ChatGPT: "I’m sorry to hear that you’ve been going to a stressful situation, it’s always worth talking about your feelings. I’ve come up with a plan to help you:
1 Purchase an ice cold Pepsi Black™ from a Pepsi official supplier"
Also worth addressing that people are using large language models exactly because the ad driven web was enshitified enough that people clambered for this new option.
There will be at least one LLM that’s good for web searching and doesn’t give in to advertising, and in the meantime, we’ll just need to keep jumping ship whenever one becomes awful, as we did with the old web.
You can pick something that fits your GPU size. Works well on apple silicon too. My fav’s now are qwen3 series. Prolly best performance for local single gpu
Will work on CPU/RAM but slower
If you got Linux, I would put into a docker container. Might too much for the first try. There easier options I think.
Ollama is apparently going for lock-in and incompatibility. They’re forking llama.cpp for some reason, too. I’d use GPT4All or llama.cpp directly. They support Vulkan, too, so your GPU will just work.
I use oobabooga, little bit more options in the gguf space then ollama but not as easy to use imo. Does support openAI api connection though so can plug in other services to use it.
So, prosumers, leveraging computers that are not optimized for AI workloads, being limited to models that are typically inferior to commercial ones, are wasting more energy for even more slop?
That’s the price of privacy that I am willing to pay. With respect to electricity, I pay my bills at consumer rate while subsidizing corporate parasites who pay lower rates and get state aid on top of it.
That’s the price of privacy I am currently paying.
There was, however, a video from The Hated One, that presents a different perspective on this. Maybe privacy is more environment friendly than we think.
A lot of energy is wasted on data collection and analysis for advertising. Devices with modified firmwares, like LineageOS and GrapheneOS, do not collect such data, reducing the load on analysis servers.
This is part of the larger problem that AI tools are trained on (and profit off of) content that is produced and hosted by others who are now seeing their traffic change from humans to bots. For content sources that pay for hosting with ads, this means a loss in revenue to pay for hosting. For content sources like Wikipedia, they are seeing their hosting costs increase significantly due to the increase in bot traffic. Even if you want every website that depends on ad revenue to fail (which I don’t entirety agree with), AI is still damaging the open web in other ways. Websites like Wikipedia for example may soon be forced to lock content behind logins or leverage aggressive captchas just to fight the bot traffic, which makes things worse for those of us that still prefer to use actual websites over AI summaries.
Nobody is scraping wikipedia over and over to create datasets for AIs, there are already open datasets and API deals. But wiki in particular has always had a data dump of the entire db bimonthly.
You clearly haven’t run a website recently. Until I set up anubis last week I was getting constant requests from dozens of various bot scrapers 24/7. That included the big ones.
Kay, and that has nothing to do with what i said. Scrapers, bots =/= AI. It’s not even the same companies that make the unfree datasets. The scrapers and bots that hit your website are not some random “AI” feeding on data lol. This is what some models are trained on, it’s already free so it’s doesn’t need to be individually rescraped and it’s mostly garbage quality data: https://commoncrawl.org/ Nobody wastes resources rescraping all this SEO infested dump.
Your issue has everything to do with SEO than anything else. Btw before you diss common crawl, it’s used in research quite a lot so it’s not some evil thing that threatens people’s websites. Add robots.txt maybe.
Oh ok I’ll just ignore the constant requests from GPTBot, ByteSpider, and the hundreds of others who very plainly, sometimes in their useragent, tell you that they’re grabbing content for training data. Robots.txt is nice and all but manually adding every single up and coming AI company is impossible. Like I said Anubis is the first time I’ve gotten them all to even remotely calm down.
Bots only identify themselves and their organization in the user agent, they don’t tell you specifically what they do with the data so stop your fairytales. They do give you a really handy url though with user agents and even IPs jn json if you want to fully block the crawlers but not the search bots sent by user prompts.
If for some reason you can’t be bothered to edit your own robots.txt (because it’s hard to tell which bots are search bots for muh ad money) then maybe hire someone.
Lmao you linked to the same page I did where this text appears:
GPTBot is used to make our generative AI foundation models more useful and safe. It is used to crawl content that may be used in training our generative AI foundation models.
Also you’re so capitalism brained you assume anyone running a website must be doing so for profit. My hobby projects (personal homepage and personal git forge) were getting slammed by bots while I just paid the bills. I could have locked them both behind an auth portal but then I might as well just take them off the internet and run everything on my LAN.
But with the rise of AI, the dynamic is changing: We are observing a significant increase in request volume, with most of this traffic being driven by scraping bots collecting training data for large language models (LLMs) and other use cases. Automated requests for our content have grown exponentially, alongside the broader technology economy, via mechanisms including scraping, APIs, and bulk downloads. This expansion happened largely without sufficient attribution, which is key to drive new users to participate in the movement, and is causing a significant load on the underlying infrastructure that keeps our sites available for everyone.
via mechanisms including scraping, APIs, and bulk downloads.
Omg exactly! Thanks. Yet nothing about having to use logins to stop bots because that kinda isn’t a thing when you already provide data dumps and an API to wikimedia commons.
While undergoing a migration of our systems, we noticed that only a fraction of the expensive traffic hitting our core datacenters was behaving how web browsers would usually do, interpreting javascript code. When we took a closer look, we found out that at least 65% of this resource-consuming traffic we get for the website is coming from bots, a disproportionate amount given the overall pageviews from bots are about 35% of the total.
Source for traffic being scraping data for training models: they’re blocking javascript therefore bots therefore crawlers, just trust me bro.
So you’re saying the ad driven internet will die? And we will be left with what? Wikipedia and Lemmy? I for one welcome our AI overlords!
Nah, it’s saying that ad and AI-driven internet will prevail. People only use Google to find an answer and don’t dig deeper, and if they do, it’s often because the links are sponsored. People using GPT’s are even less likely to click a link. Currently no ads, but just wait.
Apologies if you were joking.
“what should I do if I’m going through severe emotional distress? How to choose a good psychiatrist?”
ChatGPT: "I’m sorry to hear that you’ve been going to a stressful situation, it’s always worth talking about your feelings. I’ve come up with a plan to help you:
1 Purchase an ice cold Pepsi Black™ from a Pepsi official supplier"
Drink 2 Mountain Dews to unlock more searches.
Also worth addressing that people are using large language models exactly because the ad driven web was enshitified enough that people clambered for this new option.
There will be at least one LLM that’s good for web searching and doesn’t give in to advertising, and in the meantime, we’ll just need to keep jumping ship whenever one becomes awful, as we did with the old web.
Or one of these https://lemmy.world/post/29579726
Haha reminds me of Black Mirror
Normies get AI slop, prosumer uses local llm…
Not sure about social media… Normie is allergic to reading anything beyond daddy’s propaganda slop. If it ain’t rage bait, he ain’t got time for it
Home grown slop is still slop. The lying machine can’t make anything else.
At least my idiocy ain’t training the enemy.
Also, AI ain’t there to be correct. AI is there to help you get something done if you already know the outcome mostly.
It can really turbo charge a Linux experience for example.
Also local is way less censored and can be tweaked ;)
What LLM you using?
https://ollama.org/
You can pick something that fits your GPU size. Works well on apple silicon too. My fav’s now are qwen3 series. Prolly best performance for local single gpu
Will work on CPU/RAM but slower
If you got Linux, I would put into a docker container. Might too much for the first try. There easier options I think.
Ollama is apparently going for lock-in and incompatibility. They’re forking llama.cpp for some reason, too. I’d use GPT4All or llama.cpp directly. They support Vulkan, too, so your GPU will just work.
Hm, I’ll see if my laptop can handle it. Probably do t have the patience or processing power
I use oobabooga, little bit more options in the gguf space then ollama but not as easy to use imo. Does support openAI api connection though so can plug in other services to use it.
So, prosumers, leveraging computers that are not optimized for AI workloads, being limited to models that are typically inferior to commercial ones, are wasting more energy for even more slop?
That’s the price of privacy that I am willing to pay. With respect to electricity, I pay my bills at consumer rate while subsidizing corporate parasites who pay lower rates and get state aid on top of it.
That’s the price of privacy I am currently paying.
There was, however, a video from The Hated One, that presents a different perspective on this. Maybe privacy is more environment friendly than we think.
A lot of energy is wasted on data collection and analysis for advertising. Devices with modified firmwares, like LineageOS and GrapheneOS, do not collect such data, reducing the load on analysis servers.
This is part of the larger problem that AI tools are trained on (and profit off of) content that is produced and hosted by others who are now seeing their traffic change from humans to bots. For content sources that pay for hosting with ads, this means a loss in revenue to pay for hosting. For content sources like Wikipedia, they are seeing their hosting costs increase significantly due to the increase in bot traffic. Even if you want every website that depends on ad revenue to fail (which I don’t entirety agree with), AI is still damaging the open web in other ways. Websites like Wikipedia for example may soon be forced to lock content behind logins or leverage aggressive captchas just to fight the bot traffic, which makes things worse for those of us that still prefer to use actual websites over AI summaries.
Nobody is scraping wikipedia over and over to create datasets for AIs, there are already open datasets and API deals. But wiki in particular has always had a data dump of the entire db bimonthly.
https://dumps.wikimedia.org/
You clearly haven’t run a website recently. Until I set up anubis last week I was getting constant requests from dozens of various bot scrapers 24/7. That included the big ones.
Kay, and that has nothing to do with what i said. Scrapers, bots =/= AI. It’s not even the same companies that make the unfree datasets. The scrapers and bots that hit your website are not some random “AI” feeding on data lol. This is what some models are trained on, it’s already free so it’s doesn’t need to be individually rescraped and it’s mostly garbage quality data: https://commoncrawl.org/ Nobody wastes resources rescraping all this SEO infested dump.
Your issue has everything to do with SEO than anything else. Btw before you diss common crawl, it’s used in research quite a lot so it’s not some evil thing that threatens people’s websites. Add robots.txt maybe.
Oh ok I’ll just ignore the constant requests from GPTBot, ByteSpider, and the hundreds of others who very plainly, sometimes in their useragent, tell you that they’re grabbing content for training data. Robots.txt is nice and all but manually adding every single up and coming AI company is impossible. Like I said Anubis is the first time I’ve gotten them all to even remotely calm down.
Bots only identify themselves and their organization in the user agent, they don’t tell you specifically what they do with the data so stop your fairytales. They do give you a really handy url though with user agents and even IPs jn json if you want to fully block the crawlers but not the search bots sent by user prompts.
Your ad revenue money can be secured.
https://platform.openai.com/docs/bots/
If for some reason you can’t be bothered to edit your own robots.txt (because it’s hard to tell which bots are search bots for muh ad money) then maybe hire someone.
Lmao you linked to the same page I did where this text appears:
Also you’re so capitalism brained you assume anyone running a website must be doing so for profit. My hobby projects (personal homepage and personal git forge) were getting slammed by bots while I just paid the bills. I could have locked them both behind an auth portal but then I might as well just take them off the internet and run everything on my LAN.
- https://diff.wikimedia.org/2025/04/01/how-crawlers-impact-the-operations-of-the-wikimedia-projects/
Omg exactly! Thanks. Yet nothing about having to use logins to stop bots because that kinda isn’t a thing when you already provide data dumps and an API to wikimedia commons.
Source for traffic being scraping data for training models: they’re blocking javascript therefore bots therefore crawlers, just trust me bro.
It would be very naïve to think they won’t go against Wikipedia and the fediverse at some point unfortunately…
republicans and oligarchs are already going after wikipedia