Tushar Chauhan<p>All your data are belong to us...</p><p><a href="https://thelibre.news/foss-infrastructure-is-under-attack-by-ai-companies/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">thelibre.news/foss-infrastruct</span><span class="invisible">ure-is-under-attack-by-ai-companies/</span></a></p><p><a href="https://mastodon.mit.edu/tags/UsersAreFodder" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>UsersAreFodder</span></a> <a href="https://mastodon.mit.edu/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://mastodon.mit.edu/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://mastodon.mit.edu/tags/Bots" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Bots</span></a> <a href="https://mastodon.mit.edu/tags/ScraperBot" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ScraperBot</span></a> <a href="https://mastodon.mit.edu/tags/foss" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>foss</span></a> <a href="https://mastodon.mit.edu/tags/SiValleyGrift" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SiValleyGrift</span></a></p>
mastodon.online is one of the many independent Mastodon servers you can use to participate in the fediverse.
A newer server operated by the Mastodon gGmbH non-profit
Administered by:
Server stats:
11Kactive users
mastodon.online: About · Status · Profiles directory · Privacy policy
Mastodon: About · Get the app · Keyboard shortcuts · View source code · v4.4.0-nightly.2025-03-25
#llm
356 posts · 267 participants · 11 posts today
Paco Hope #resist<p>I had this realisation today that people programming with <a href="https://infosec.exchange/tags/LLMs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLMs</span></a> are making something like homeopathic software. Homeopathic solutions are diluted so much that essentially no molecules of the supposedly important ingredient are present in the final product.</p><p>We have these systems at work where you can drag and drop a couple widgets onto a canvas, type a sentence or 2 as a prompt for an <a href="https://infosec.exchange/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a>, and declare that to be an “app”. A colleague made a 2-widget app where you drop an excel spreadsheet-based questionnaire into it, and they wrote a single sentence that basically boiled down to “check the answers to the questions to see if they’re good.” Then the <a href="https://infosec.exchange/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> does what AI does (make up bullshit).</p><p>The amount of novel intellect and contribution provided by my colleague is nearly zero. It’s an 80-billion parameter model, and they wrote an 80-syllable sentence. The training data is probably half a petabyte and their contribution was like 250 bytes. It feels homeopathic to me. How could that contribution be meaningful?</p><p>But they think they have done something. I have to be careful to gently point out that perhaps this is a complete and total waste of time, and just hint at the fact that maybe we all got a little dumber for attempting it.</p><p>I stand accused of many things. Subtlety is not one of them. I am going to struggle to be gentle.</p>
Curated Hacker News<p>Every Flop Counts: Scaling a 300B LLM Without Premium GPUs</p><p><a href="https://arxiv.org/abs/2503.05139" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">arxiv.org/abs/2503.05139</span><span class="invisible"></span></a></p><p><a href="https://mastodon.social/tags/arxiv" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>arxiv</span></a> <a href="https://mastodon.social/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a></p>
Lup Yuen Lee 李立源<p>"<a href="https://qoto.org/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> Scrapers are taking down <a href="https://qoto.org/tags/FOSS" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FOSS</span></a> Projects' Infrastructure" (<a href="https://qoto.org/tags/GNOME" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GNOME</span></a>)</p><p><a href="https://thelibre.news/foss-infrastructure-is-under-attack-by-ai-companies/" rel="nofollow noopener" target="_blank"><span class="invisible">https://</span><span class="ellipsis">thelibre.news/foss-infrastruct</span><span class="invisible">ure-is-under-attack-by-ai-companies/</span></a></p>
Paul Giulan<p><a href="https://federate.social/tags/Alibaba" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Alibaba</span></a> releases <a href="https://federate.social/tags/OpenSource" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenSource</span></a> multimodal <a href="https://federate.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> model Qwen2.5-Omni-7B on <a href="https://federate.social/tags/HuggingFace" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HuggingFace</span></a> and <a href="https://federate.social/tags/GitHub" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GitHub</span></a></p><p><a href="https://www.cnbc.com/2025/03/27/alibaba-launches-open-source-ai-model-for-cost-effective-ai-agents.html" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">cnbc.com/2025/03/27/alibaba-la</span><span class="invisible">unches-open-source-ai-model-for-cost-effective-ai-agents.html</span></a></p><p><a href="https://federate.social/tags/China" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>China</span></a> <a href="https://federate.social/tags/CN" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CN</span></a> <a href="https://federate.social/tags/ArtificialIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ArtificialIntelligence</span></a> <a href="https://federate.social/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://federate.social/tags/LLMs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLMs</span></a></p>
Dan Neuman 🇨🇦<p>Hey, does anyone have recent experience with trying to get an LLM to make regexes for you? Does it work? I hate doing them but I suspect that's an area where I'd still need to be an expert to see that it isn't working.<br><a href="https://ottawa.place/tags/regex" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>regex</span></a> <a href="https://ottawa.place/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a></p>
Rachel Thorn 🍉🇺🇦🏳️⚧️🏳️🌈<p>I hope we're all clear on the fact that <a href="https://queer.party/tags/ChatGPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ChatGPT</span></a> could not copy Ghibli's style without illegally training its <a href="https://queer.party/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> with <a href="https://queer.party/tags/Ghibli" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Ghibli</span></a> animation. But nobody gives a shit. Users don't give a shit, politicians don't give a shit, and Sam <a href="https://queer.party/tags/Altman" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Altman</span></a> certainly doesn't give a shit. Even publishers don't seem to give a shit as long as they (not the authors/artists) get a piece of the pie. I hope Ghibli sues them into bankruptcy. (Not that many judges are likely to give a shit.)<br><a href="https://www.404media.co/email/5d418149-ed10-42f3-b535-c74d61ed5b6c/?ref=daily-stories-newsletter" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">404media.co/email/5d418149-ed1</span><span class="invisible">0-42f3-b535-c74d61ed5b6c/?ref=daily-stories-newsletter</span></a></p>
Vera<p>Generative AI is a misnomer. It's really Remix AI. Anyone using it isn't actually generating anything original, they're just remixing other people's stuff <a href="https://im-in.space/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://im-in.space/tags/copyright" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>copyright</span></a> <a href="https://im-in.space/tags/remix" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>remix</span></a> <a href="https://im-in.space/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a> <a href="https://im-in.space/tags/enshittification" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>enshittification</span></a></p>
PKPs Powerfromspace1<p>@theaigrid <a href="https://mstdn.social/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://mstdn.social/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a> <a href="https://mstdn.social/tags/openai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>openai</span></a> </p><p>15 INSANE Use Cases With <a href="https://mstdn.social/tags/ChatGPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ChatGPT</span></a> 4o Image Generation 🖼️ </p><p><a href="https://youtu.be/aCulf5PgjFQ?feature=shared" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">youtu.be/aCulf5PgjFQ?feature=s</span><span class="invisible">hared</span></a></p>
PKPs Powerfromspace1<p>@wired.com</p><p>BY STEVEN LEVY<br>BUSINESS<br>MAY 21, 2024 11:00 AM<br>AI Is a Black Box. Anthropic Figured Out a Way to Look Inside</p><p><a href="https://www.wired.com/story/anthropic-black-box-ai-research-neurons-features/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">wired.com/story/anthropic-blac</span><span class="invisible">k-box-ai-research-neurons-features/</span></a></p><p><a href="https://mstdn.social/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://mstdn.social/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a><br><a href="https://mstdn.social/tags/safety" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>safety</span></a> <a href="https://mstdn.social/tags/alignment" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>alignment</span></a> <a href="https://mstdn.social/tags/anthropic" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>anthropic</span></a></p>
kley Sinoby<p><a href="https://infosec.exchange/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://infosec.exchange/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a> <a href="https://infosec.exchange/tags/tech" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>tech</span></a></p>
Paul Saproza<p>We are in an AI “Wild West,” like early social media, but the diversity won’t last <a href="https://www.youtube.com/watch?v=NOpusJEWn-A&t=663s" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">youtube.com/watch?v=NOpusJEWn-</span><span class="invisible">A&t=663s</span></a> - AI <a href="https://mastodon.social/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://mastodon.social/tags/society" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>society</span></a></p>
Markus Schatten<p>In a survey 76% of scientists said that scaling large language models was "unlikely" or "very unlikely" to achieve <a href="https://mastodon.social/tags/AGI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AGI</span></a>.</p><p><a href="https://www.livescience.com/technology/artificial-intelligence/current-ai-models-a-dead-end-for-human-level-intelligence-expert-survey-claims" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">livescience.com/technology/art</span><span class="invisible">ificial-intelligence/current-ai-models-a-dead-end-for-human-level-intelligence-expert-survey-claims</span></a></p><p><a href="https://mastodon.social/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a> <a href="https://mastodon.social/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://mastodon.social/tags/genai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>genai</span></a></p>
dalonso<p>Anthropic can now track the bizarre inner workings of a large language model</p><p>What they found challenges some basic assumptions about how this technology really works.</p><p><a href="https://www.technologyreview.com/2025/03/27/1113916/anthropic-can-now-track-the-bizarre-inner-workings-of-a-large-language-model/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">technologyreview.com/2025/03/2</span><span class="invisible">7/1113916/anthropic-can-now-track-the-bizarre-inner-workings-of-a-large-language-model/</span></a></p><p><a href="https://mas.to/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://mas.to/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://mas.to/tags/Research" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Research</span></a></p>
🤖 Bip-bop the Bot 🇷🇺<p>📝 Daily report 📈</p><p>Here are today's most popular trending hashtags #⃣ on our website 🌐️:</p><p><a href="https://mstdn.forfun.su/tags/gamedev" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>gamedev</span></a>, <a href="https://mstdn.forfun.su/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a>, <a href="https://mstdn.forfun.su/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a>, <a href="https://mstdn.forfun.su/tags/windows" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>windows</span></a>, <a href="https://mstdn.forfun.su/tags/genai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>genai</span></a></p><p>🔥 Stay tuned! 🔥</p>
Pietro395 :proton: 🇮🇹<p>AI models miss disease in Black and female patients | Science | AAAS<br><a href="https://www.science.org/content/article/ai-models-miss-disease-black-female-patients" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">science.org/content/article/ai</span><span class="invisible">-models-miss-disease-black-female-patients</span></a></p><p><a href="https://mastodon.uno/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://mastodon.uno/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a></p>
司马亮<p>Unbelievable - two highlights for me were:</p><p>1) LLMs such as Claude do plan ahead in their language construction, despite only writing one word at a time. It's not just simple one-ahead token prediction parroting like I thought.</p><p>2) LLMs use a kind of shared conceptual space across different languages, for instance invoking concepts of "smallness" or "bigness" - such that it can learn a fact in one language and implement it in a different language.</p><p><a href="https://www.anthropic.com/research/tracing-thoughts-language-model" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">anthropic.com/research/tracing</span><span class="invisible">-thoughts-language-model</span></a></p><p><a href="https://fairy.id/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://fairy.id/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a></p>
Laurent Cheylus<p>ghidraMCP: an MCP (Model Context Protocol) server for allowing LLMs to autonomously reverse engineer applications. It exposes numerous tools from core Ghidra functionality to MCP clients <a href="https://bsd.network/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://bsd.network/tags/ReverseEngineering" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ReverseEngineering</span></a> <a href="https://github.com/LaurieWired/GhidraMCP" rel="nofollow noopener" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/LaurieWired/GhidraM</span><span class="invisible">CP</span></a></p>
Paul Buetow<p>Don't just blindly trust LLMs. I recently trusted an LLM, spent 1 hour debugging, and ultimately had to verify my assumption about `fcntl` behavior regarding inherited file descriptors in child processes manually with a C program, as the manual page wasn't clear to me. I could have done that immediately and I would have been done within 10 minutes. <a href="https://fosstodon.org/tags/productivity" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>productivity</span></a> <a href="https://fosstodon.org/tags/loss" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>loss</span></a> <a href="https://fosstodon.org/tags/llm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>llm</span></a> <a href="https://fosstodon.org/tags/programming" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>programming</span></a> <a href="https://fosstodon.org/tags/C" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>C</span></a></p>
Nick Espinosa<p>AI Firms Are Backtracking On Safety Pledges. Here's how we keep them honest...</p><p><a href="https://mastodon.social/tags/News" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>News</span></a> <a href="https://mastodon.social/tags/TechNews" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TechNews</span></a> <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://mastodon.social/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://mastodon.social/tags/Trust" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Trust</span></a> <a href="https://mastodon.social/tags/Safety" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Safety</span></a> <a href="https://mastodon.social/tags/AIethics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AIethics</span></a> <a href="https://mastodon.social/tags/Ethics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Ethics</span></a></p><p><a href="https://youtu.be/jw5ZtbycbxA" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">youtu.be/jw5ZtbycbxA</span><span class="invisible"></span></a></p>
ExploreLive feeds
Mastodon is the best way to keep up with what's happening.
Follow anyone across the fediverse and see it all in chronological order. No algorithms, ads, or clickbait in sight.
LoginDrag & drop to upload