<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: Icko</title><link>https://news.ycombinator.com/user?id=Icko</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 05 Jun 2026 03:01:32 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=Icko" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by Icko in "They’re made out of weights"]]></title><description><![CDATA[
<p>Good thing that is not poetry</p>
]]></description><pubDate>Thu, 04 Jun 2026 10:01:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=48396400</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=48396400</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48396400</guid></item><item><title><![CDATA[New comment by Icko in "72% of the dollar's purchasing power was destroyed in just four episodes"]]></title><description><![CDATA[
<p>I saw an article that went really into depth - basically oil has huge diversity, and they really wanted Venezuelan oil:<p><a href="https://open.substack.com/pub/endtropy/p/trumps-enormous-c-length-win-over" rel="nofollow">https://open.substack.com/pub/endtropy/p/trumps-enormous-c-l...</a></p>
]]></description><pubDate>Mon, 30 Mar 2026 16:28:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=47576372</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=47576372</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47576372</guid></item><item><title><![CDATA[New comment by Icko in "DBRX: A new open LLM"]]></title><description><![CDATA[
<p>I'm 90% certain that OpenAI has some much beefier model they are not releasing - remember the Q* rumour?</p>
]]></description><pubDate>Wed, 27 Mar 2024 14:33:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=39839702</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=39839702</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39839702</guid></item><item><title><![CDATA[New comment by Icko in "Dijkstra's interview on Dutch TV (2000)"]]></title><description><![CDATA[
<p>I feel you should be able to keep the high level idea in your head at once; of course not the details. I don't think I've ever completed a piece of code without having a scheme in my head, and where it should fit in it.</p>
]]></description><pubDate>Mon, 05 Feb 2024 00:52:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=39256134</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=39256134</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39256134</guid></item><item><title><![CDATA[New comment by Icko in "Transformers Are Multi-State RNNs"]]></title><description><![CDATA[
<p>I've seen at least 6 such papers, all being like "<popular architecture> are actually <a bit older concept>". Neural networks are generic enough that you can make them equivalent to almost everything.</p>
]]></description><pubDate>Tue, 16 Jan 2024 20:51:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=39018963</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=39018963</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39018963</guid></item><item><title><![CDATA[New comment by Icko in "Eating fewer calories can ward off ageing"]]></title><description><![CDATA[
<p>I can't get a read on this guy. I have strong bullshit feelings, but I'm not 100% sure.<p>1. He seems to have an awful lot of referral links - $60 for a bottle of olive oil?
2. He has not posted <i>why</i> is his regime like that - e.g. "Here is a paper about iron deficiencies, so I'm taking an iron supplement". 
3. 90% of his website is "look at how dense my bones are", "look at how healthy my liver is" - bro I just wanted to see the actual regime, and you can't even post that.</p>
]]></description><pubDate>Tue, 02 Jan 2024 08:38:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=38839474</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=38839474</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38839474</guid></item><item><title><![CDATA[New comment by Icko in "Predictive policing software terrible at predicting crimes"]]></title><description><![CDATA[
<p>No, it would still be a crime. He wouldn't know to pursue it, but it is very much a crime.</p>
]]></description><pubDate>Wed, 04 Oct 2023 10:10:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=37763045</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=37763045</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37763045</guid></item><item><title><![CDATA[New comment by Icko in "Predictive policing software terrible at predicting crimes"]]></title><description><![CDATA[
<p>None of those are victimless crimes. If you attempt to shoot someone and your rifle jams, you still exposed him to the risk of dying. This is a ridiculous strawman.</p>
]]></description><pubDate>Wed, 04 Oct 2023 10:08:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=37763036</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=37763036</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37763036</guid></item><item><title><![CDATA[New comment by Icko in "A Multi-Level View of LLM Intentionality"]]></title><description><![CDATA[
<p>> Primarily, they are pure functions that accept a sequence of tokens and return the next token. The model itself is stateless, and it doesn't seem right to me to ascribe "intent" to a stateless function. Even if the function is capable of modeling certain aspects of chess.<p>I have two arguments against. One, you could argue that state is transferred between the layers. It may be inelegant for each chain of state transitions to be the same length, but it seems to work. Two, it may not have "states", but if the end result is the same, does it matter?</p>
]]></description><pubDate>Mon, 11 Sep 2023 20:33:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=37473288</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=37473288</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37473288</guid></item><item><title><![CDATA[New comment by Icko in "The Arab Dagger"]]></title><description><![CDATA[
<p>How are the scabbards so curvy? Is the blade simply very short?</p>
]]></description><pubDate>Sun, 27 Aug 2023 00:16:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=37278134</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=37278134</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37278134</guid></item><item><title><![CDATA[New comment by Icko in "Llama: Add grammar-based sampling"]]></title><description><![CDATA[
<p>How is this different from Guidance and LMQL?</p>
]]></description><pubDate>Fri, 21 Jul 2023 23:27:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=36821404</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=36821404</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36821404</guid></item><item><title><![CDATA[New comment by Icko in "OpenAI isn’t doing enough to make ChatGPT’s limitations clear"]]></title><description><![CDATA[
<p>That's a pretty good idea, actually. It'll be a bitch and a half to tune the threshold so it's reliable, and doesn't mark normally generated text though.</p>
]]></description><pubDate>Tue, 30 May 2023 19:00:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=36129068</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=36129068</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36129068</guid></item><item><title><![CDATA[New comment by Icko in "Robot Landlords Are Buying Up Houses"]]></title><description><![CDATA[
<p>As far as I know, all adversarial attacks need to know the model... But that is a very interesting line of thought.</p>
]]></description><pubDate>Tue, 29 Nov 2022 15:54:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=33789093</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=33789093</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33789093</guid></item><item><title><![CDATA[New comment by Icko in "Frog: OCR Tool for Linux"]]></title><description><![CDATA[
<p>There was "EAST OCR" detector, which is basically someone put a deep learning model in openCV somehow. <a href="https://www.folio3.ai/blog/text-detection-by-using-opencv-and-east/" rel="nofollow">https://www.folio3.ai/blog/text-detection-by-using-opencv-an...</a></p>
]]></description><pubDate>Thu, 24 Nov 2022 16:03:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=33732843</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=33732843</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33732843</guid></item><item><title><![CDATA[New comment by Icko in "Frog: OCR Tool for Linux"]]></title><description><![CDATA[
<p>Last I compared them, (1-2 years ago), Google OCR was much much better and supported more languages than tesseract. There was also an OCR in openCV, which was slightly better than tesseract, but not good enough to be useful.</p>
]]></description><pubDate>Tue, 22 Nov 2022 12:17:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=33704953</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=33704953</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33704953</guid></item><item><title><![CDATA[New comment by Icko in "Ask HN: What are your favorite books/courses of 2022? Why?"]]></title><description><![CDATA[
<p>I went through Delta academy's Reinforcement Learning courses. I really like having someone to ask questions when I got stuck. Not free, but 100% worth the money.<p><a href="https://joindeltaacademy.com/" rel="nofollow">https://joindeltaacademy.com/</a></p>
]]></description><pubDate>Mon, 21 Nov 2022 09:06:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=33690452</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=33690452</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33690452</guid></item><item><title><![CDATA[New comment by Icko in "Ask HN: Tips for Reducing the Number of Inboxes"]]></title><description><![CDATA[
<p>We work with lots of russian/ukrainian devs, and we're using Telegram.</p>
]]></description><pubDate>Mon, 23 May 2022 11:27:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=31477643</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=31477643</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=31477643</guid></item><item><title><![CDATA[New comment by Icko in "Show HN: Oldest Search – Search for the oldest result on internet"]]></title><description><![CDATA[
<p>Haha, time definitely proved you right!</p>
]]></description><pubDate>Wed, 11 May 2022 12:12:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=31338751</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=31338751</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=31338751</guid></item><item><title><![CDATA[New comment by Icko in "Mental speed is high until age sixty"]]></title><description><![CDATA[
<p>perhaps reading speed?</p>
]]></description><pubDate>Mon, 18 Apr 2022 10:52:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=31069211</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=31069211</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=31069211</guid></item><item><title><![CDATA[New comment by Icko in "FBI Didn't Knock Down a Suspect's Door Because 'It Was an Affluent Neighborhood'"]]></title><description><![CDATA[
<p>run for office, you seem to know what to do</p>
]]></description><pubDate>Mon, 11 Apr 2022 06:18:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=30985308</link><dc:creator>Icko</dc:creator><comments>https://news.ycombinator.com/item?id=30985308</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=30985308</guid></item></channel></rss>