Artificial intelligence: Difference between revisions

(4 intermediate revisions by 2 users not shown)

Line 62:

</blockquote>

====MediaWiki====

====MediaWiki, Wikipedia, and the Wikimedia Foundation====

[[wikipedia:MediaWiki|MediaWiki]] is of particular interest to LLM training due to the vast amount of factual, plain-text content wikis tend to hold. While [[wikipedia:Wikipedia|Wikipedia]] and the [[wikipedia:Wikimedia Foundation|Wikimedia Foundation]] host the most well-known wikis, numerous smaller wikis exist thanks to the work of many independent editors. The strength of wiki architecture is its ability for every edit to be audited by anyone, at any time - you can still view [https://en.wikipedia.org/w/index.php?oldid=1 the first edit to Wikipedia] from 2002. This makes wikis a hybrid of a static website and a dynamic web app, which becomes problematic when poorly-designed bots attempt to scrape them.<ref name="geraspora" />

<!-- COI alert: I, [[User:kirb]], am an admin for The Apple Wiki. Hopefully this is neutral enough?

-->The Apple Wiki, which documents internal details of Apple's hardware and software, holds more than 50,000 articles. On 2 August 2024, with a repeat occurrence on 5 January 2025, the service was disrupted by scraping efforts.<ref>https://theapplewiki.com/wiki/The_Apple_Wiki:Community_portal#Bot_traffic_abuse</ref> The wiki contains a considerable amount of information that is scraped by legitimate security research tools, making it difficult for the website to block non-legitimate requests. Efforts to block unethical scraping and protect the wiki have disrupted these legitimate tools. The large article count, combined with more than 280,000 total edits over the wiki's lifetime, create an untenable situation where it is simply not possible to scrape the website without causing significant service disruption.

On 1 April 2025, the Wikimedia Foundation indicated that its infrastructure has been under increasing pressure from content scraping bots since January 2024, with the particularly critical metric that "65% of our most expensive traffic comes from bots", despite estimating 35% of all traffic as coming from bots. The bots create traffic patterns that are significantly unlike human traffic patterns, effectively bypassing Wikimedia's caching infrastructure and placing significant load on the core servers. A blog post provides an example where bot traffic caused the [[wikipedia:Wikimedia Commons|Wikimedia Commons]] service to become unstable during a human traffic spike. The Foundation is considering introduction of a Responsible Use of Infrastructure policy to ensure the continued stability of their services.<ref>https://diff.wikimedia.org/2025/04/01/how-crawlers-impact-the-operations-of-the-wikimedia-projects/</ref>

====Perplexity AI and news outlets====

[[Perplexity AI]], founded in August 2022, is a large language model that aims to be viewed as a general search engine. It encourages users to consume news through its summaries of stories.

On 15 June 2024, Apple blog MacStories found that Perplexity does not follow its own documented policies when accessing content the user requests from the web. In their testing, the scraper pretended to be Chrome 111 running on Windows 10, connecting from an IP address not found in Perplexity's ~~posted~~ IP address ranges.<ref>https://rknight.me/blog/perplexity-ai-is-lying-about-its-user-agent/</ref> ~~Two days later, this was corroborated~~ by WIRED.<ref>https://www.wired.com/story/perplexity-is-a-bullshit-machine/</ref> Perplexity responded by removing its list of IP addresses.

On 15 June 2024, an investigation by Apple blog MacStories found that Perplexity does not follow its own documented policies when accessing content the user requests from the web. In their testing, the scraper pretended to be Chrome 111 running on Windows 10, connecting from an IP address not found in Perplexity's publicly-listed IP address ranges.<ref>https://rknight.me/blog/perplexity-ai-is-lying-about-its-user-agent/</ref> MacStories' findings were confirmed by a WIRED investigation.<ref>https://www.wired.com/story/perplexity-is-a-bullshit-machine/</ref> Perplexity responded by removing its list of IP addresses.

On 27 June 2024, [[Amazon]] announced an investigation into Perplexity AI, ~~citing a~~ terms of service ~~clause requiring bots hosted on Amazon Web Services to honor robots.txt~~:<ref name="perplexity-aws">https://www.wired.com/story/aws-perplexity-bot-scraping-investigation/</ref>

On 27 June 2024, [[Amazon]] announced an investigation into Perplexity AI, suggesting the behavior may be considered abusive under Amazon Web Services terms of service:<ref name="perplexity-aws">https://www.wired.com/story/aws-perplexity-bot-scraping-investigation/</ref>

Line 79:

Line 81:

</blockquote>

== ~~Privacy concerns of online AI models~~ ==

====Read the Docs====

~~There are several concerns with using online~~ AI ~~models like [[ChatGPT]]~~, ~~not only because they are proprietary~~, ~~but also because there is no guarantee to where your data ends up being stored or used for~~.

In an early example, on 25 July 2024, open source documentation website Read the Docs detailed cases of abusive bots downloading large amounts of content from the service. Particularly, the significant range of IP addresses used in an aggressive manner rendered existing rate limiting ineffective. Taking action to block traffic identified by Cloudflare as "AI crawlers" reduced bandwidth requirements by 75%, at a cost saving of $1,500 USD/month.<ref>https://about.readthedocs.com/blog/2024/07/ai-crawlers-abuse/</ref>

~~Luckily there is~~ an alternative which solves many of these concerns, which is to run AI models locally. There currently exist different models that are small enough to run on a personal computer. Those models are indicated with a smaller parameter size, for ~~instance models with 1~~.~~5B or 7B parameters~~. If the ~~computer has a relatively modern GPU, it can also run one~~ of ~~the larger models~~ for ~~more accurate answers~~, ~~as these models have GPU-acceleration~~. ~~The software that will be recommended below runs on all major computer platforms (Windows~~/~~macOs~~/~~Linux)~~. ~~Be cautious if you download other kinds of models besides the major models, as platforms like HuggingFace allow anyone to upload~~.<ref>https://~~huggingface~~.co/</ref>

====SourceHut and Fedora Linux====

On 15 March 2025, an infrastructure manager for the [[wikipedia:Fedora Linux|Fedora Linux]] open source project discussed an assumed large language model crawling attack against the Prague.io Git source code hosting service. The project made the decision to block the entire country of Brazil for some time, while also blocking access to certain repositories whose traffic was creating significant CPU usage.<ref>https://www.scrye.com/blogs/nirik/posts/2025/03/15/mid-march-infra-bits-2025/</ref><ref>https://www.scrye.com/blogs/nirik/posts/2025/03/29/late-march-infra-bits-2025/</ref>

~~=== LM Studio ===~~

On 17 March 2025, the Git source code host SourceHut announced that the service was being disrupted by large language model crawlers. Mitigations deployed to reduce disruption involved requiring login for some areas of the service, and blocking IP ranges of cloud providers, affecting legitimate use of the website by its users.<ref>https://status.sr.ht/issues/2025-03-17-git.sr.ht-llms/</ref> In response to the event, SourceHut founder Drew DeVault wrote a blog post entitled "[https://drewdevault.com/2025/03/17/2025-03-17-Stop-externalizing-your-costs-on-me.html Please stop externalizing your costs directly into my face]", discussing his frustrations with having ongoing and ever-adapting attacks that must be addressed in a timely fashion to reduce disruption to legitimate SourceHut users. DeVault estimates that between "20-100%" of his time is now spent addressing such attacks.

~~One~~ of the ~~easiest software to start with to run these models is LM Studio~~.<ref>https://~~lmstudio~~.ai/</ref> ~~It is user-friendly as it has a graphical user interface aimed at beginners, and allows you~~ to ~~get started with just a few clicks. It recommends appropriately sized models for your specific computer hardware, and manages the rest of~~ the ~~installation for you. In terms of storage~~, ~~you will need~~ a ~~few gigabytes to store the models locally, which you only have to do once. With the models installed, no further internet connection is required.<ref>~~[https://~~www.youtube~~.com/~~@NetworkChuck NetworkChuck]: [https:~~//~~www.youtube.com~~/~~watch?v=7TR~~-~~FLWNVHY The only way to run deepseek~~]~~</ref> The software allows opening chats~~ with ~~the large language model, which you can also organize into folders~~.

==~~= Ollama =~~==

==Privacy concerns of online AI models==

~~If you~~ are ~~fine~~ with ~~just~~ using ~~the terminal~~, ~~another option~~ is to ~~install software~~ like ~~Ollama~~.<ref>https://~~ollama~~.~~com~~/</ref> ~~Once installed, you can simply invoke the run command with the model you want to use, and it will download that model if it has not done that already. The website lists the most common~~ models to run, like Llama ([[Meta]]) DeepSeek ([[DeepSeek]]), Phi ([[Microsoft]]), Mistral ([[Mistral AI]]), Gemma ([[Google]]). If you are a more advanced user, you can also run Ollama inside Docker.<ref>https://ollama.com/blog/ollama-is-now-available-as-an-official-docker-image</ref> That allows isolating the model completely from your host system, which may be what you want to be extra secure.

There are several concerns with using online AI models like [[ChatGPT]] ([[OpenAI]]), not only because they are proprietary, but also because there is no guarantee to where your data ends up being stored or used for. Recent developments in local AI models are an alternative to these online AI models, as they work offline once they are downloaded from platforms like HuggingFace.<ref>https://huggingface.co/</ref> Common models to run are like Llama ([[Meta]]), DeepSeek ([[DeepSeek]]), Phi ([[Microsoft]]), Mistral ([[Mistral AI]]), Gemma ([[Google]]).

==References==

@@ Line 62: / Line 62: @@
 </blockquote>
-====MediaWiki====
+====MediaWiki, Wikipedia, and the Wikimedia Foundation====
 [[wikipedia:MediaWiki|MediaWiki]] is of particular interest to LLM training due to the vast amount of factual, plain-text content wikis tend to hold. While [[wikipedia:Wikipedia|Wikipedia]] and the [[wikipedia:Wikimedia Foundation|Wikimedia Foundation]] host the most well-known wikis, numerous smaller wikis exist thanks to the work of many independent editors. The strength of wiki architecture is its ability for every edit to be audited by anyone, at any time - you can still view [https://en.wikipedia.org/w/index.php?oldid=1 the first edit to Wikipedia] from 2002. This makes wikis a hybrid of a static website and a dynamic web app, which becomes problematic when poorly-designed bots attempt to scrape them.<ref name="geraspora" />
 <!-- COI alert: I, [[User:kirb]], am an admin for The Apple Wiki. Hopefully this is neutral enough?
 -->The Apple Wiki, which documents internal details of Apple's hardware and software, holds more than 50,000 articles. On 2 August 2024, with a repeat occurrence on 5 January 2025, the service was disrupted by scraping efforts.<ref>https://theapplewiki.com/wiki/The_Apple_Wiki:Community_portal#Bot_traffic_abuse</ref> The wiki contains a considerable amount of information that is scraped by legitimate security research tools, making it difficult for the website to block non-legitimate requests. Efforts to block unethical scraping and protect the wiki have disrupted these legitimate tools. The large article count, combined with more than 280,000 total edits over the wiki's lifetime, create an untenable situation where it is simply not possible to scrape the website without causing significant service disruption.
+On 1 April 2025, the Wikimedia Foundation indicated that its infrastructure has been under increasing pressure from content scraping bots since January 2024, with the particularly critical metric that "65% of our most expensive traffic comes from bots", despite estimating 35% of all traffic as coming from bots. The bots create traffic patterns that are significantly unlike human traffic patterns, effectively bypassing Wikimedia's caching infrastructure and placing significant load on the core servers. A blog post provides an example where bot traffic caused the [[wikipedia:Wikimedia Commons|Wikimedia Commons]] service to become unstable during a human traffic spike. The Foundation is considering introduction of a Responsible Use of Infrastructure policy to ensure the continued stability of their services.<ref>https://diff.wikimedia.org/2025/04/01/how-crawlers-impact-the-operations-of-the-wikimedia-projects/</ref>
 ====Perplexity AI and news outlets====
 [[Perplexity AI]], founded in August 2022, is a large language model that aims to be viewed as a general search engine. It encourages users to consume news through its summaries of stories.
-On 15 June 2024, Apple blog MacStories found that Perplexity does not follow its own documented policies when accessing content the user requests from the web. In their testing, the scraper pretended to be Chrome 111 running on Windows 10, connecting from an IP address not found in Perplexity's posted IP address ranges.<ref>https://rknight.me/blog/perplexity-ai-is-lying-about-its-user-agent/</ref> Two days later, this was corroborated by WIRED.<ref>https://www.wired.com/story/perplexity-is-a-bullshit-machine/</ref> Perplexity responded by removing its list of IP addresses.
+On 15 June 2024, an investigation by Apple blog MacStories found that Perplexity does not follow its own documented policies when accessing content the user requests from the web. In their testing, the scraper pretended to be Chrome 111 running on Windows 10, connecting from an IP address not found in Perplexity's publicly-listed IP address ranges.<ref>https://rknight.me/blog/perplexity-ai-is-lying-about-its-user-agent/</ref> MacStories' findings were confirmed by a WIRED investigation.<ref>https://www.wired.com/story/perplexity-is-a-bullshit-machine/</ref> Perplexity responded by removing its list of IP addresses.
-On 27 June 2024, [[Amazon]] announced an investigation into Perplexity AI, citing a terms of service clause requiring bots hosted on Amazon Web Services to honor robots.txt:<ref name="perplexity-aws">https://www.wired.com/story/aws-perplexity-bot-scraping-investigation/</ref>
+On 27 June 2024, [[Amazon]] announced an investigation into Perplexity AI, suggesting the behavior may be considered abusive under Amazon Web Services terms of service:<ref name="perplexity-aws">https://www.wired.com/story/aws-perplexity-bot-scraping-investigation/</ref>
 <blockquote>
@@ Line 79: / Line 81: @@
 </blockquote>
-== Privacy concerns of online AI models ==
+====Read the Docs====
-There are several concerns with using online AI models like [[ChatGPT]], not only because they are proprietary, but also because there is no guarantee to where your data ends up being stored or used for.
+In an early example, on 25 July 2024, open source documentation website Read the Docs detailed cases of abusive bots downloading large amounts of content from the service. Particularly, the significant range of IP addresses used in an aggressive manner rendered existing rate limiting ineffective. Taking action to block traffic identified by Cloudflare as "AI crawlers" reduced bandwidth requirements by 75%, at a cost saving of $1,500 USD/month.<ref>https://about.readthedocs.com/blog/2024/07/ai-crawlers-abuse/</ref>
-Luckily there is an alternative which solves many of these concerns, which is to run AI models locally. There currently exist different models that are small enough to run on a personal computer. Those models are indicated with a smaller parameter size, for instance models with 1.5B or 7B parameters. If the computer has a relatively modern GPU, it can also run one of the larger models for more accurate answers, as these models have GPU-acceleration. The software that will be recommended below runs on all major computer platforms (Windows/macOs/Linux). Be cautious if you download other kinds of models besides the major models, as platforms like HuggingFace allow anyone to upload.<ref>https://huggingface.co/</ref>
+====SourceHut and Fedora Linux====
+On 15 March 2025, an infrastructure manager for the [[wikipedia:Fedora Linux|Fedora Linux]] open source project discussed an assumed large language model crawling attack against the Prague.io Git source code hosting service. The project made the decision to block the entire country of Brazil for some time, while also blocking access to certain repositories whose traffic was creating significant CPU usage.<ref>https://www.scrye.com/blogs/nirik/posts/2025/03/15/mid-march-infra-bits-2025/</ref><ref>https://www.scrye.com/blogs/nirik/posts/2025/03/29/late-march-infra-bits-2025/</ref>
-=== LM Studio ===
+On 17 March 2025, the Git source code host SourceHut announced that the service was being disrupted by large language model crawlers. Mitigations deployed to reduce disruption involved requiring login for some areas of the service, and blocking IP ranges of cloud providers, affecting legitimate use of the website by its users.<ref>https://status.sr.ht/issues/2025-03-17-git.sr.ht-llms/</ref> In response to the event, SourceHut founder Drew DeVault wrote a blog post entitled "[https://drewdevault.com/2025/03/17/2025-03-17-Stop-externalizing-your-costs-on-me.html Please stop externalizing your costs directly into my face]", discussing his frustrations with having ongoing and ever-adapting attacks that must be addressed in a timely fashion to reduce disruption to legitimate SourceHut users. DeVault estimates that between "20-100%" of his time is now spent addressing such attacks.
-One of the easiest software to start with to run these models is LM Studio.<ref>https://lmstudio.ai/</ref> It is user-friendly as it has a graphical user interface aimed at beginners, and allows you to get started with just a few clicks. It recommends appropriately sized models for your specific computer hardware, and manages the rest of the installation for you. In terms of storage, you will need a few gigabytes to store the models locally, which you only have to do once. With the models installed, no further internet connection is required.<ref>[https://www.youtube.com/@NetworkChuck NetworkChuck]: [https://www.youtube.com/watch?v=7TR-FLWNVHY The only way to run deepseek]</ref> The software allows opening chats with the large language model, which you can also organize into folders.
-=== Ollama ===
+==Privacy concerns of online AI models==
-If you are fine with just using the terminal, another option is to install software like Ollama.<ref>https://ollama.com/</ref> Once installed, you can simply invoke the run command with the model you want to use, and it will download that model if it has not done that already. The website lists the most common models to run, like Llama ([[Meta]]) DeepSeek ([[DeepSeek]]), Phi ([[Microsoft]]), Mistral ([[Mistral AI]]), Gemma ([[Google]]). If you are a more advanced user, you can also run Ollama inside Docker.<ref>https://ollama.com/blog/ollama-is-now-available-as-an-official-docker-image</ref> That allows isolating the model completely from your host system, which may be what you want to be extra secure.
+There are several concerns with using online AI models like [[ChatGPT]] ([[OpenAI]]), not only because they are proprietary, but also because there is no guarantee to where your data ends up being stored or used for. Recent developments in local AI models are an alternative to these online AI models, as they work offline once they are downloaded from platforms like HuggingFace.<ref>https://huggingface.co/</ref> Common models to run are like Llama ([[Meta]]), DeepSeek ([[DeepSeek]]), Phi ([[Microsoft]]), Mistral ([[Mistral AI]]), Gemma ([[Google]]).
 ==References==