<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: tripzilch</title><link>https://news.ycombinator.com/user?id=tripzilch</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 10 Jun 2026 12:06:07 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=tripzilch" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by tripzilch in "You can no longer Google the word 'disregard'"]]></title><description><![CDATA[
<p>Today I learn :)</p>
]]></description><pubDate>Sat, 23 May 2026 23:10:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=48252515</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=48252515</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48252515</guid></item><item><title><![CDATA[New comment by tripzilch in "You can no longer Google the word 'disregard'"]]></title><description><![CDATA[
<p>From what I can tell, this isn't actually a response/protection against prompt-injection (which is what I imagined from the article's title).<p>It seems to be just the AI responding to the word "disregard".<p>I tried searching for the phrase "never mind" (like, the Nirvana album), and Google's AI responded similarly:<p><pre><code>    AI Overview
    never mind
    No problem at all! Just let me know if there's anything else I can help you with later. Have a great day!
</code></pre>
It's just a terrible user experience because the AI misinterpreted the search query, and the actual web search results are pushed "below the fold", on purpose.</p>
]]></description><pubDate>Sat, 23 May 2026 17:40:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=48249537</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=48249537</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48249537</guid></item><item><title><![CDATA[New comment by tripzilch in "Frontier AI has broken the open CTF format"]]></title><description><![CDATA[
<p>I don't understand the complainiture, it's an improve</p>
]]></description><pubDate>Sun, 17 May 2026 10:03:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=48167507</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=48167507</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48167507</guid></item><item><title><![CDATA[New comment by tripzilch in "Clay PCB Tutorial"]]></title><description><![CDATA[
<p>I love that idea (the "anti-masterwork")</p>
]]></description><pubDate>Mon, 27 Apr 2026 13:29:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=47921315</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47921315</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47921315</guid></item><item><title><![CDATA[New comment by tripzilch in "The AI industry is discovering that the public hates it"]]></title><description><![CDATA[
<p>There's also the part where a lot of AIs were trained on data people did not consent to be used.</p>
]]></description><pubDate>Mon, 27 Apr 2026 12:39:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=47920825</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47920825</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47920825</guid></item><item><title><![CDATA[New comment by tripzilch in "An update on recent Claude Code quality reports"]]></title><description><![CDATA[
<p>I don't think it's fair or reasonable to charge <i>your</i> cache misses to the user.</p>
]]></description><pubDate>Sat, 25 Apr 2026 08:40:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=47899785</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47899785</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47899785</guid></item><item><title><![CDATA[New comment by tripzilch in "Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model"]]></title><description><![CDATA[
<p>I believe that if you pay them for their time, it's not really "wasted", at least not nearly as "wasted" as when the next person would pay them to design some vapid advertisement.<p>In addition to that 1) it's for science and 2) maybe you owe it to yourself to have a really nice framed picture of a pelican riding a bicycle on the wall :D<p>About the dedicated image generation results, I still would have made the bicycle smaller, but it starts to depend on how motivated the artist is to make both the bike and pelican accurate. Which is fine, but if you want to have a benchmark, it's important to have at least one "known good" example, I think.</p>
]]></description><pubDate>Fri, 24 Apr 2026 17:13:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=47893089</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47893089</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47893089</guid></item><item><title><![CDATA[New comment by tripzilch in "Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model"]]></title><description><![CDATA[
<p>IMHO looks more like a stork, not a pelican. Look up any image of an actual pelican and check the ratio of legs to body. IMHO that's a weird mistake to make when asked for a "pelican".<p>Have you considered asking a couple of artists on Fiverr or something to draw you a picture with the same prompt? I don't mean this as a gotcha, it's actual advice, you should probably get a sense of what a real human artist/designer (or three) would do with this prompt.<p>For example, I hope you will find that: One <i>reasoning choice</i> is wrong with this picture that's not much to do with its ability to draw. Do we enlarge the pelican to human size? Or do we shrink the bike to pelican size? There is only one answer that keeps pelican proportions. Draw a pelican on a very tiny bike, and its legs will just fit without making it a different species, and you can even sort of cover part of the steer under the wings, etc etc.<p>I'm curious if other artists would come up with the same or other solutions, but they should in general come up with <i>solutions</i>, which I haven't seen the LLM do, really.<p>You (or maybe others?) said that the "pelican on a bike" prompt is good because "there is no right answer" cause you can't really fit a pelican on a bike. But most artists will say "hold my beer" and figure it out anyway. Cartoonists won't even have to think. The "figuring out" of these problems is what I'm missing in the LLMs response. It just put a pelican on a bike and makes it look like a stork if necessary. I don't really feel like it's actually testing for the thing this prompt is designed for, unless the test <i>still</i> says "FAIL" for each and all of them, including the one you just called "excellent".</p>
]]></description><pubDate>Thu, 23 Apr 2026 12:33:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=47875004</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47875004</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47875004</guid></item><item><title><![CDATA[New comment by tripzilch in "Meta to start capturing employee mouse movements, keystrokes for AI training"]]></title><description><![CDATA[
<p>But surely you could imagine that things work differently in different places.</p>
]]></description><pubDate>Wed, 22 Apr 2026 15:45:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=47865270</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47865270</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47865270</guid></item><item><title><![CDATA[New comment by tripzilch in "Qwen3.6-Plus: Towards real world agents"]]></title><description><![CDATA[
<p>I hate to be the one to tell you this but things changed rather quickly when they elected a clown dictator, and now the US is widely considered to be a low-key geopolitical rival to the 'developed West in general' (blergh) including Canada and Europe ...<p>Seriously, even if you manage to elect someone capable ever again, the US can't be trusted to not elect someone worse than a toddler again in 4 years. In fact, you can't even be trusted to not elect someone even <i>worse</i> than your current dictator (if you thought it couldn't get worse, Trump isn't the bottom of the barrel <i>by far</i>).</p>
]]></description><pubDate>Sat, 04 Apr 2026 08:53:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=47637234</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47637234</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47637234</guid></item><item><title><![CDATA[New comment by tripzilch in "The bot situation on the internet is worse than you could imagine"]]></title><description><![CDATA[
<p>>  I hate ads, and I hate having to serve ads. We get some subscription users but nowhere near enough to cover costs.<p>I hate ads and I hate having to use an ad blocker to be able to not go crazy in order to use the Internet.<p>You merely hate "having" to serve ads because it denies you profit from the people you're exploiting with those ads.<p>Why is your business more deserving to exist on the Internet than my usage??</p>
]]></description><pubDate>Mon, 30 Mar 2026 08:53:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=47572018</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47572018</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47572018</guid></item><item><title><![CDATA[New comment by tripzilch in "Vatican Rebukes Peter Thiel's Antichrist Lectures in Rome"]]></title><description><![CDATA[
<p>he named his company "Palantir" after <i>Lord of the Rings</i> fantasy books and didn't get it exactly right either</p>
]]></description><pubDate>Mon, 23 Mar 2026 19:45:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=47494223</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47494223</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47494223</guid></item><item><title><![CDATA[New comment by tripzilch in "Vatican Rebukes Peter Thiel's Antichrist Lectures in Rome"]]></title><description><![CDATA[
<p>> One simple argument for changing form is that the Sun will eventually scorch Earth so we do have an expiry date.<p>uhhhh this is utterly ridiculous<p>the sun will expire and make the Earth uninhabitable in ~1B years<p>humans have existed in their current form for 300K years<p>so that's about one three-millionth of that time<p>> You might hate the guy but he's not dumb<p>you really have to be a special kind of dumb to believe humanity will snuff it because of something that'll not happen until 3.33 MILLION times as long as humans have walked on this Earth<p>that just makes no sense however you turn it, there is not a single thing that this has been the case for in the (much much longer) history of Earth, let alone humanity<p>And I wasn't talking about Peter Thiel being a special kind of dumb. If he's so smart he probably also secretly laughs in the face of people who go along with this reasoning.</p>
]]></description><pubDate>Mon, 23 Mar 2026 19:41:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=47494170</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47494170</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47494170</guid></item><item><title><![CDATA[New comment by tripzilch in "Warranty Void If Regenerated"]]></title><description><![CDATA[
<p>No please, I also agree with parent poster. Talk to the LLM, cause the human ain't listening.</p>
]]></description><pubDate>Thu, 19 Mar 2026 12:22:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=47438124</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47438124</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47438124</guid></item><item><title><![CDATA[New comment by tripzilch in "Malus – Clean Room as a Service"]]></title><description><![CDATA[
<p>What about my latest algorithm, VibeSort<p><pre><code>    // VibeSort
    let arr = [51,46,72,32,14,27,88,32];

    arr.sort((a,b)=>{
      let response = LLM.query(`Which number is larger, number A:${a} or number B:${b}. Answer using "A" or "B" only, if they are equal, say "C".`);
      if(response.includes('C')) return 0;
      if(response.includes('B')) return -1;
      if(response.includes('A')) return 1;
      return 0;
    });

    console.table(arr);</code></pre></p>
]]></description><pubDate>Fri, 13 Mar 2026 10:42:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=47362671</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47362671</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47362671</guid></item><item><title><![CDATA[New comment by tripzilch in "Don't post generated/AI-edited comments. HN is for conversation between humans"]]></title><description><![CDATA[
<p>Trying to keep things on topic, BTW, I found that LLMs are pretty good at picking up the kinds of context that makes this very obvious what is really being meant.<p>So you could use an LLM, privately, to soften people's opinions.<p>I just tried it for you, I won't copy it here cause the thread is about not using LLMs, but if you get too upset from somebody being simply direct and clear in their manner of speaking, the LLM is trained on enough American cultural baggage that it is very capable of softening that blow with the extra words you so dearly need to see past that red mist.<p>Someone might even be able to vibe code a browser plugin for it.</p>
]]></description><pubDate>Thu, 12 Mar 2026 11:54:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=47349368</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47349368</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47349368</guid></item><item><title><![CDATA[New comment by tripzilch in "Don't post generated/AI-edited comments. HN is for conversation between humans"]]></title><description><![CDATA[
<p>It's completely clear what is intended, the only thing you're disagreeing about, is the cultural difference of who is expected to make this translation.<p>I think that would've been pretty clear from the post too, if you weren't so keen on giving a non-native speaker an English lesson ...</p>
]]></description><pubDate>Thu, 12 Mar 2026 11:45:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=47349306</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47349306</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47349306</guid></item><item><title><![CDATA[New comment by tripzilch in "The whole thing was a scam"]]></title><description><![CDATA[
<p>> Simply spending money to get someone you like elected isn’t bribery.<p>Alright then what should it be called, because it's also not democracy.</p>
]]></description><pubDate>Sun, 01 Mar 2026 13:40:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=47206608</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47206608</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47206608</guid></item><item><title><![CDATA[New comment by tripzilch in "I am directing the Department of War to designate Anthropic a supply-chain risk"]]></title><description><![CDATA[
<p>> Everyone running Claude on AWS, boom gone. The level of disruption to the US economy would be off the charts<p>seriously? :)</p>
]]></description><pubDate>Sat, 28 Feb 2026 10:12:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=47193167</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47193167</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47193167</guid></item><item><title><![CDATA[New comment by tripzilch in "US orders diplomats to fight data sovereignty initiatives"]]></title><description><![CDATA[
<p>The reason why we're not keeping lists of which people believe what religion, is because such lists were extremely useful to the nazis in WW2 when exterminating Jewish people.<p>> Think Orthodox Jews<p>Pretty sure they would remember why this is the case.<p>> Regulating a sacred scroll like a database table seems wrong.<p>There is actually no perceivable or material difference between something that is considered "sacred" and that which is not. It really hinges on whether some subset of some splinter of some religion considers it so.<p>But, I'm not familiar with these cases you mention, I think there's some details left out that should matter. The really weird thing to me, is that a sports club can keep a list of members easily (yes they need to abide by the GDPR but it's not hard), and if somehow a "religious group" can't manage that level of organization, I don't think their opinion on what objects are considered "sacred" should count for much, either.<p>Another issue is that "religious groups" can have a different opinion of who are their members and who they get to keep data on, and it doesn't matter whether those records are "sacred" or not, according to the GDPR it is not the "religious group", but the people whose data is being kept whose opinion counts. It would be ridiculous otherwise. I had to email a Church to stop tracking me (which happens if you're baptized as a baby), and that should be my choice, it would be insane if they could claim "yeah tough luck, but our records are sacred".</p>
]]></description><pubDate>Thu, 26 Feb 2026 12:59:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=47165454</link><dc:creator>tripzilch</dc:creator><comments>https://news.ycombinator.com/item?id=47165454</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47165454</guid></item></channel></rss>