<p>The future of communication in two panels. <br><a href="/tags/ai/" rel="tag">#AI</a></p>
ai
<p>I've been reading up on the Lottery Ticket Hypothesis, which is super interesting.</p><p>Basically, the observation is that these days we build vast neural networks with billions of parameters, but most of the parameters aren't needed. That is, after training, you can just throw away 95% of the network (pruning), and it will still work fine.</p><p>The LTH paper is asking: could we start with a network just 5% of the size, and get comparable results? If so, that would be a huge performance win for Deep Learning.</p><p>What's interesting is that you can do this, but only by training the full network (perhaps several times) to see which neurons are needed. They argue that training a neural network isn't so much creating a model, as finding a lucky sub-network (a lottery ticket) from the randomly initialized network, a bit like a sculpter "finding" the bust hidden in a block of marble.</p><p>Initial LTH paper: <a href="http://arxiv.org/abs/1803.03635" rel="nofollow"><span class="invisible">http://</span>arxiv.org/abs/1803.03635</a><br>Follow-up with major clarifications: <a href="http://arxiv.org/abs/1905.01067" rel="nofollow"><span class="invisible">http://</span>arxiv.org/abs/1905.01067</a></p><p><a href="/tags/science/" rel="tag">#science</a> <a href="/tags/ai/" rel="tag">#ai</a> <a href="/tags/machinelearning/" rel="tag">#machinelearning</a></p>
<p>Oh wow, someone reporting losing a "year of Ableton projects" after running an AI-generated script.</p><p>"I didn’t realize what it was doing and ran it without backing anything up."</p><p>There's a few lessons to be learned here.</p><p><a href="https://www.reddit.com/r/ableton/comments/1kmnhfu/lost_a_year_of_ableton_projects_after_running_a/" rel="nofollow" class="ellipsis" title="www.reddit.com/r/ableton/comments/1kmnhfu/lost_a_year_of_ableton_projects_after_running_a/"><span class="invisible">https://</span><span class="ellipsis">www.reddit.com/r/ableton/comme</span><span class="invisible">nts/1kmnhfu/lost_a_year_of_ableton_projects_after_running_a/</span></a></p><p><a href="/tags/musicians/" rel="tag">#musicians</a> <a href="/tags/backup/" rel="tag">#backup</a> <a href="/tags/ai/" rel="tag">#AI</a></p>
<p>"Wait, isn't this the plot to the "Terminator" movies?"</p><p><a href="https://futurism.com/openai-signs-deal-us-government-nuclear-weapon-security" rel="nofollow" class="ellipsis" title="futurism.com/openai-signs-deal-us-government-nuclear-weapon-security"><span class="invisible">https://</span><span class="ellipsis">futurism.com/openai-signs-deal</span><span class="invisible">-us-government-nuclear-weapon-security</span></a></p><p><a href="/tags/news/" rel="tag">#news</a> <a href="/tags/technology/" rel="tag">#technology</a> <a href="/tags/technews/" rel="tag">#TechNews</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/openai/" rel="tag">#OpenAI</a> <a href="/tags/military/" rel="tag">#military</a> <a href="/tags/terminator/" rel="tag">#terminator</a></p>
<p>ai, didn't read.<br>brilliant.</p><p><a href="/tags/artificialintelligence/" rel="tag">#artificialintelligence</a> <a href="/tags/ai/" rel="tag">#ai</a></p>
<p>There is no AI, just other people's data.</p><p><a href="/tags/robot/" rel="tag">#robot</a> <a href="/tags/comic/" rel="tag">#comic</a> <a href="/tags/ai/" rel="tag">#ai</a> <a href="/tags/data/" rel="tag">#data</a> <a href="/tags/people/" rel="tag">#people</a> <a href="/tags/noai/" rel="tag">#noai</a></p>
Edited 1y ago
<p>Did you know? <a href="/tags/fedify/" rel="tag">#Fedify</a> provides <a href="/tags/documentation/" rel="tag">#documentation</a> optimized for LLMs through the <a href="https://llmstxt.org/" rel="nofollow">llms.txt standard</a>.</p><p>Available endpoints:</p><p><a href="https://fedify.dev/llms.txt" rel="nofollow"><span class="invisible">https://</span>fedify.dev/llms.txt</a> — Core documentation overview<br><a href="https://fedify.dev/llms-full.txt" rel="nofollow"><span class="invisible">https://</span>fedify.dev/llms-full.txt</a> — Complete documentation dump</p><p>Useful for training <a href="/tags/ai/" rel="tag">#AI</a> assistants on <a href="/tags/activitypub/" rel="tag">#ActivityPub</a>/<a href="/tags/fediverse/" rel="tag">#fediverse</a> development, building documentation chatbots, or <a href="/tags/llm/" rel="tag">#LLM</a>-powered dev tools.</p><p><a href="/tags/fedidev/" rel="tag">#fedidev</a></p>
If you turn the sink off when you're done using it to conserve water, or turn the lights off when you leave the room to conserve electricity, why do you use ChatGPT or other AI tools? Using those sorts of tools a few times a month negates whatever you've conserved by being prudent in other parts of your life.<br><br><a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/genai/" rel="tag">#GenAI</a> <a href="/tags/generativeai/" rel="tag">#GenerativeAI</a> <a href="/tags/llm/" rel="tag">#LLM</a> <a href="/tags/waste/" rel="tag">#waste</a> <a href="/tags/environment/" rel="tag">#environment</a><br>
<p>The <a href="/tags/eu/" rel="tag">#EU</a> bans some <a href="/tags/ai/" rel="tag">#AI</a> uses and regulates most of the rest, in service of its citizens and not the "techbros".</p><p>AI systems with 'unacceptable risk' are now banned in the EU | TechCrunch<br><a href="https://techcrunch.com/2025/02/02/ai-systems-with-unacceptable-risk-are-now-banned-in-the-eu/" rel="nofollow" class="ellipsis" title="techcrunch.com/2025/02/02/ai-systems-with-unacceptable-risk-are-now-banned-in-the-eu/"><span class="invisible">https://</span><span class="ellipsis">techcrunch.com/2025/02/02/ai-s</span><span class="invisible">ystems-with-unacceptable-risk-are-now-banned-in-the-eu/</span></a></p>
<p>I think one of, if not *the* main purpose of <a href="/tags/ai/" rel="tag">#AI</a> is that anything can be claimed as fake or faked by the powers that be. This makes it an excellent tool of <a href="/tags/fascism/" rel="tag">#fascism</a>. And the more it's normalized, *for any reason* the better for them. It's the potential for them to twist and to reshape reality. <a href="/tags/fuckai/" rel="tag">#FuckAI</a>, <a href="/tags/fuckbillionaires/" rel="tag">#FuckBillionaires</a>, <a href="/tags/fuckfascists/" rel="tag">#FuckFascists</a>, and fuck their "post-truth" world.</p><p>Do not give in. Resist. Dare to be inconvenienced, dare to do things the "hard way".</p>
Edited 148d ago
<p>Resistance to the coup is the defense of the human against the digital and the democratic against the oligarchic.<br></p>From <a href="https://snyder.substack.com/p/of-course-its-a-coup" rel="nofollow" class="ellipsis" title="snyder.substack.com/p/of-course-its-a-coup"><span class="invisible">https://</span><span class="ellipsis">snyder.substack.com/p/of-cours</span><span class="invisible">e-its-a-coup</span></a><br><br>Defense of the human against the digital has been my mission for some time. Resisting the narratives about how <a href="/tags/llms/" rel="tag">#LLMs</a> "reason", "pass the Turing test", "diagnose illnesses", are "better than humans" in various ways are part of it. Resisting the false narrative that we're on the verge of discovering <a href="/tags/agi/" rel="tag">#AGI</a> is part of it. Allowing these false stories to persist and spread means succumbing to very dark anti-human forces. We're seeing some of the consequences now, and we're seeing how far this might go.<br><br><a href="/tags/uspol/" rel="tag">#USPol</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/genai/" rel="tag">#GenAI</a> <a href="/tags/generativeai/" rel="tag">#GenerativeAI</a> <a href="/tags/llm/" rel="tag">#LLM</a> <a href="/tags/agi/" rel="tag">#AGI</a><br>
<p>Some argue that ai technology is more significant than electricity or the internet, and so it will spread fast. But there is little sign of this. Only 5-6% of American businesses said they used ai to produce goods and services in 2024, according to the country’s Census Bureau.<br></p>From <a href="https://www.economist.com/catch-up-rampage-in-new-orleans-tesla-cybertruck-explosion/2024/12/31/the-ai-productivity-puzzle" rel="nofollow" class="ellipsis" title="www.economist.com/catch-up-rampage-in-new-orleans-tesla-cybertruck-explosion/2024/12/31/the-ai-productivity-puzzle"><span class="invisible">https://</span><span class="ellipsis">www.economist.com/catch-up-ram</span><span class="invisible">page-in-new-orleans-tesla-cybertruck-explosion/2024/12/31/the-ai-productivity-puzzle</span></a> , titled The AI productivity puzzle<br><br>What's the puzzle? It has its uses but generally this technology is not particularly useful for most people. Countless billions of dollars have been spent hyping it up to pretend that isn't the case. But reality matters.<br><br>Nowadays I think of generative AI as a power grab. It's not a useful tool for a lot of people, but it's exceptionally useful to people who want to grab and hold power, or restructure power in their favor.<br><br><a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/genai/" rel="tag">#GenAI</a> <a href="/tags/generativeai/" rel="tag">#GenerativeAI</a><br>
<p>"The original video showing off Google's Gemini offerings in Workspace was posted five days ago and showed the AI generating a response for a small-business owner in Wisconsin. In the response, Gemini claims Gouda accounts for "50 to 60% of the world's cheese consumption.</p><p>The edited ad now shows a Gemini response that says Gouda "is one of the most popular cheeses in the world."</p><p><a href="https://www.businessinsider.com/google-super-bowl-ad-gemini-ai-gouda-cheese-2025-2" rel="nofollow" class="ellipsis" title="www.businessinsider.com/google-super-bowl-ad-gemini-ai-gouda-cheese-2025-2"><span class="invisible">https://</span><span class="ellipsis">www.businessinsider.com/google</span><span class="invisible">-super-bowl-ad-gemini-ai-gouda-cheese-2025-2</span></a></p><p>Archived link: <a href="https://archive.is/2cX3i" rel="nofollow"><span class="invisible">https://</span>archive.is/2cX3i</a></p><p><a href="/tags/news/" rel="tag">#news</a> <a href="/tags/technews/" rel="tag">#TechNews</a> <a href="/tags/technology/" rel="tag">#technology</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/google/" rel="tag">#google</a> <a href="/tags/ads/" rel="tag">#ads</a></p>
<p>Useful <a href="/tags/ai/" rel="tag">#AI</a> acronyms:</p><p><a href="/tags/aicrap/" rel="tag">#aiCRAP</a> - Computer-Rendered Artificial Pictures<br><a href="/tags/aipiss/" rel="tag">#aiPISS</a> - Plagiarized Information Synthesis System<br><a href="/tags/aiarse/" rel="tag">#aiARSE</a> - Auto-Regurgitated Syntactic Extrusion</p>
Edited 41d ago
<p>Not sure what the chances of this bill passing are, but it's certainly interesting to see the efforts.</p><p>"Under Hawley’s proposed law, “technology or intellectual property” developed in China would be barred from importation into the U.S. Anyone found violating these restrictions could face up to 20 years in prison, as well as substantial monetary penalties of up to $1 million for individuals and $100 million for companies."</p><p><a href="https://interestingengineering.com/culture/law-proposed-to-criminalize-deepseek" rel="nofollow" class="ellipsis" title="interestingengineering.com/culture/law-proposed-to-criminalize-deepseek"><span class="invisible">https://</span><span class="ellipsis">interestingengineering.com/cul</span><span class="invisible">ture/law-proposed-to-criminalize-deepseek</span></a></p><p><a href="/tags/news/" rel="tag">#news</a> <a href="/tags/technews/" rel="tag">#TechNews</a> <a href="/tags/technology/" rel="tag">#technology</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/deepseek/" rel="tag">#deepseek</a></p>
<p>Unbelievable</p><p><a href="/tags/elonmusk/" rel="tag">#ElonMusk</a>’s US <a href="/tags/doge/" rel="tag">#DOGE</a> Service are feeding sensitive data into <a href="/tags/ai/" rel="tag">#AI</a> software via <a href="/tags/microsoft/" rel="tag">#Microsoft</a>’s <a href="/tags/cloud/" rel="tag">#cloud</a></p><p><a href="/tags/musk/" rel="tag">#Musk</a>’s US <a href="/tags/doge/" rel="tag">#DOGE</a> Service have fed sensitive data from across the <a href="/tags/education/" rel="tag">#Education</a> Dept into <a href="/tags/artificialintelligence/" rel="tag">#ArtificialIntelligence</a> software to probe the agency’s programs & spending….
The AI probe includes data w/personally identifiable info for people who manage grants, & sensitive internal financial data…</p><p><a href="/tags/law/" rel="tag">#law</a> <a href="/tags/security/" rel="tag">#security</a> <a href="/tags/infosec/" rel="tag">#InfoSec</a> <a href="/tags/cybersecurity/" rel="tag">#CyberSecurity</a> <a href="/tags/nationalsecurity/" rel="tag">#NationalSecurity</a> <a href="/tags/trump/" rel="tag">#Trump</a> <a href="/tags/trumpcoup/" rel="tag">#TrumpCoup</a><br><a href="https://www.washingtonpost.com/nation/2025/02/06/elon-musk-doge-ai-department-education/" rel="nofollow" class="ellipsis" title="www.washingtonpost.com/nation/2025/02/06/elon-musk-doge-ai-department-education/"><span class="invisible">https://</span><span class="ellipsis">www.washingtonpost.com/nation/</span><span class="invisible">2025/02/06/elon-musk-doge-ai-department-education/</span></a></p>
<p>"Inevitably creating the potential for bots emailing other bots and absolutely nothing else to worry about."</p><p><a href="https://www.pcgamer.com/software/ai/deepmind-boss-would-pay-thousands-of-dollars-per-month-to-get-rid-of-his-email-so-google-is-working-on-a-next-gen-gmail-ai-that-will-answer-them-in-your-style-and-maybe-make-some-of-the-easier-decisions-for-you/" rel="nofollow" class="ellipsis" title="www.pcgamer.com/software/ai/deepmind-boss-would-pay-thousands-of-dollars-per-month-to-get-rid-of-his-email-so-google-is-working-on-a-next-gen-gmail-ai-that-will-answer-them-in-your-style-and-maybe-make-some-of-the-easier-decisions-for-you/"><span class="invisible">https://</span><span class="ellipsis">www.pcgamer.com/software/ai/de</span><span class="invisible">epmind-boss-would-pay-thousands-of-dollars-per-month-to-get-rid-of-his-email-so-google-is-working-on-a-next-gen-gmail-ai-that-will-answer-them-in-your-style-and-maybe-make-some-of-the-easier-decisions-for-you/</span></a></p><p><a href="/tags/news/" rel="tag">#news</a> <a href="/tags/technology/" rel="tag">#technology</a> <a href="/tags/technews/" rel="tag">#TechNews</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/deadinternet/" rel="tag">#DeadInternet</a> <a href="/tags/email/" rel="tag">#email</a> <a href="/tags/bots/" rel="tag">#bots</a></p>
<p>On the limits of LLMs (Large Language models) and LRMs (Large Reasoning Models). The TL;DR: "Our findings reveal fundamental limitations in current models: despite sophisticated self-reflection mechanisms, these models fail to develop generalizable reasoning capabilities beyond certain complexity thresholds." Meaning: accuracy collapse.</p><p>Interesting paper from Apple. <a href="https://ml-site.cdn-apple.com/papers/the-illusion-of-thinking.pdf" rel="nofollow" class="ellipsis" title="ml-site.cdn-apple.com/papers/the-illusion-of-thinking.pdf"><span class="invisible">https://</span><span class="ellipsis">ml-site.cdn-apple.com/papers/t</span><span class="invisible">he-illusion-of-thinking.pdf</span></a></p><p><a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/llm/" rel="tag">#LLM</a> <a href="/tags/lrm/" rel="tag">#LRM</a></p>
Edited 1y ago
<p>Happy <a href="/tags/fishfriday/" rel="tag">#FishFriday</a> !</p><p>Time to <a href="/tags/makesomethingfishy/" rel="tag">#MakeSomethingFishy</a> </p><p>**not affiliated with <a href="/tags/hashtaggames/" rel="tag">#HashtagGames</a>** but my shameless promotion of my Ebay store where I sell hand crafted aquarium ornaments and other things. But feel free to play along!</p><p><a href="https://www.ebay.com/usr/crystal_fish_caves" rel="nofollow" class="ellipsis" title="www.ebay.com/usr/crystal_fish_caves"><span class="invisible">https://</span><span class="ellipsis">www.ebay.com/usr/crystal_fish_</span><span class="invisible">caves</span></a></p><p><a href="/tags/handmade/" rel="tag">#handmade</a> <a href="/tags/crafts/" rel="tag">#crafts</a> <a href="/tags/fedigiftshop/" rel="tag">#fedigiftshop</a> <a href="/tags/fishofmastodon/" rel="tag">#FishofMastodon</a> <a href="/tags/pets/" rel="tag">#pets</a> <a href="/tags/diy/" rel="tag">#diy</a> no <a href="/tags/ai/" rel="tag">#ai</a> photoshop only</p>
<p>Happy <a href="/tags/pridemonth/" rel="tag">#PrideMonth</a> <br><a href="/tags/ai/" rel="tag">#ai</a> <a href="/tags/aiart/" rel="tag">#aiart</a></p>
<p>"Australia’s Department of the Treasury has found that Microsoft’s Copilot can easily deliver return on investment, but staff exposed to the AI assistant came away from the experience less confident it will help them at work."</p><p><a href="https://www.theregister.com/2025/02/12/australian_treasury_copilot_pilot_assessment/" rel="nofollow" class="ellipsis" title="www.theregister.com/2025/02/12/australian_treasury_copilot_pilot_assessment/"><span class="invisible">https://</span><span class="ellipsis">www.theregister.com/2025/02/12</span><span class="invisible">/australian_treasury_copilot_pilot_assessment/</span></a></p><p><a href="/tags/news/" rel="tag">#news</a> <a href="/tags/technews/" rel="tag">#TechNews</a> <a href="/tags/technology/" rel="tag">#technology</a> <a href="/tags/ai/" rel="tag">#AI</a></p>
<p>We obtained records showing how a DOGE staffer with no medical experience used AI to identify which VA contracts to kill.</p><p>His tool concluded that more than a thousand deals were each worth $34 million, when in fact some were for as little as $35,000. <br><a href="https://www.propublica.org/article/trump-doge-veterans-affairs-ai-contracts-health-care?utm_source=mastodon&utm_medium=social&utm_campaign=mastodon-post" rel="nofollow" class="ellipsis" title="www.propublica.org/article/trump-doge-veterans-affairs-ai-contracts-health-care?utm_source=mastodon&utm_medium=social&utm_campaign=mastodon-post"><span class="invisible">https://</span><span class="ellipsis">www.propublica.org/article/tru</span><span class="invisible">mp-doge-veterans-affairs-ai-contracts-health-care?utm_source=mastodon&utm_medium=social&utm_campaign=mastodon-post</span></a></p><p><a href="/tags/news/" rel="tag">#News</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/doge/" rel="tag">#DOGE</a> <a href="/tags/veterans/" rel="tag">#Veterans</a> <a href="/tags/health/" rel="tag">#Health</a> <a href="/tags/healthcare/" rel="tag">#Healthcare</a> <a href="/tags/government/" rel="tag">#Government</a></p>
<p>Opt Out of AI Tracking by Disabling Google Gemini</p><p>Learn how to defend your privacy by opting out of artificial intelligence (AI) tracking by disabling Google Gemini in this comprehensive guide.</p><p><a href="https://www.adamsdesk.com/posts/disable-google-gemini/" rel="nofollow" class="ellipsis" title="www.adamsdesk.com/posts/disable-google-gemini/"><span class="invisible">https://</span><span class="ellipsis">www.adamsdesk.com/posts/disabl</span><span class="invisible">e-google-gemini/</span></a></p><p><a href="/tags/blog/" rel="tag">#blog</a> <a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/privacy/" rel="tag">#privacy</a></p>
Simulated food does not nourish you any more than simulated thought educates you.<br><br><a href="/tags/ai/" rel="tag">#AI</a> <a href="/tags/genai/" rel="tag">#GenAI</a> <a href="/tags/generativeai/" rel="tag">#GenerativeAI</a><br>
<p>Ich würde gerne <a href="/tags/opensource/" rel="tag">#opensource</a> <a href="/tags/ai/" rel="tag">#AI</a> benutzen um meine medizinischen Daten zu analysieren.</p><p><a href="https://youtu.be/ZbpnqpCifSY?si=ZE6SZYwdgF9sun-p" rel="nofollow" class="ellipsis" title="youtu.be/ZbpnqpCifSY?si=ZE6SZYwdgF9sun-p"><span class="invisible">https://</span><span class="ellipsis">youtu.be/ZbpnqpCifSY?si=ZE6SZY</span><span class="invisible">wdgF9sun-p</span></a></p>