<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: ucha</title><link>https://news.ycombinator.com/user?id=ucha</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Tue, 04 Aug 2026 00:27:20 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=ucha" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by ucha in "Qwen3.8-Max: A New Bar for Coding and Cowork"]]></title><description><![CDATA[
<p>For a set parameter size, if you benchmark an LLM on a single language, you'd get better results by training it on multiple languages rather than just the one. Scaling quality training data almost always increases performance as the neural network picks up general rules and an understanding of code that are language agnostic.</p>
]]></description><pubDate>Mon, 03 Aug 2026 14:50:23 +0000</pubDate><link>https://news.ycombinator.com/item?id=49156565</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=49156565</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49156565</guid></item><item><title><![CDATA[New comment by ucha in "Qwen3.8-Max: A New Bar for Coding and Cowork"]]></title><description><![CDATA[
<p>That's not how LLMs work. If you're talking about number of parameters, you wouldn't be able to reduce the size much by "removing" support for other languages.</p>
]]></description><pubDate>Mon, 03 Aug 2026 08:59:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=49153077</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=49153077</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49153077</guid></item><item><title><![CDATA[New comment by ucha in "Codex Resets"]]></title><description><![CDATA[
<p>Why add a middleman when you can ask your own account with /usage?</p>
]]></description><pubDate>Sun, 19 Jul 2026 09:23:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=48966348</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=48966348</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48966348</guid></item><item><title><![CDATA[Nvidia Isn't Enron – So What Is It?]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.wheresyoured.at/nvidia-isnt-enron-so-what-is-it/">https://www.wheresyoured.at/nvidia-isnt-enron-so-what-is-it/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46205465">https://news.ycombinator.com/item?id=46205465</a></p>
<p>Points: 2</p>
<p># Comments: 1</p>
]]></description><pubDate>Tue, 09 Dec 2025 14:47:53 +0000</pubDate><link>https://www.wheresyoured.at/nvidia-isnt-enron-so-what-is-it/</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=46205465</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46205465</guid></item><item><title><![CDATA[New comment by ucha in "Music eases surgery and speeds recovery, study finds"]]></title><description><![CDATA[
<p>n=56 doesn't give you much information regarding the margin of error, unless you practice Tarot Reader's cat science. The standard deviation of outcomes and the difference between both outcomes matter just as much.<p>If I flip a coin 56 times and it always falls on head, I can be pretty much certain that it's not a fair coin. I wouldn't need to flip it 1000 times. We are all someone else's "know-nothing hack"...</p>
]]></description><pubDate>Thu, 27 Nov 2025 13:01:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=46068853</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=46068853</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46068853</guid></item><item><title><![CDATA[New comment by ucha in "OpenAI dropped the price of o3 by 80%"]]></title><description><![CDATA[
<p>Can we know for sure that the price drop is accompanied by a change in the model such as quantization?<p>On twitter, some people say that some models perform better at night when there is a less demand which allows them to serve a non-quantized model.<p>Since the models are only available through API and there is no test to check which version of the model is served, it's hard to know what we're buying...</p>
]]></description><pubDate>Wed, 11 Jun 2025 18:11:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=44250242</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=44250242</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44250242</guid></item><item><title><![CDATA[New comment by ucha in "Intellect-2 Release: The First 32B Model Trained Through Globally Distributed RL"]]></title><description><![CDATA[
<p>It needs to not have economic value but it doesn't necessarily need to be useless and wasteful.</p>
]]></description><pubDate>Mon, 12 May 2025 09:48:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=43961237</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=43961237</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43961237</guid></item><item><title><![CDATA[New comment by ucha in "Growing trade deficit is selling the nation out from under us (2003) [pdf]"]]></title><description><![CDATA[
<p>You can have growing deficits without actually "selling the nation" when you account for growth and inflation.<p>Say foreigners own 50% of US assets. When someone creates in the US a company ex nihilo and sells 10% of it to foreign investors, even though foreigners bought more US assets, this new wealth creation causes the % of foreign-owned US assets to diminish.<p>When foreign investors own US bonds, as inflation increases and the money supply increases, they end up owning less in real terms.<p>This doesn't account for investments that are limited in quantity such as real estate.<p>Also just because those are opposing forces, doesn't mean that foreign ownership wouldn't increase. As a matter of fact, when looking at the foreign ownership of US stocks, it has been steadily increasing.</p>
]]></description><pubDate>Fri, 04 Apr 2025 11:20:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=43580642</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=43580642</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43580642</guid></item><item><title><![CDATA[Can LLMs make trade-offs involving stipulated pain and pleasure states?]]></title><description><![CDATA[
<p>Article URL: <a href="https://arxiv.org/abs/2411.02432">https://arxiv.org/abs/2411.02432</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=42899486">https://news.ycombinator.com/item?id=42899486</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 01 Feb 2025 16:19:03 +0000</pubDate><link>https://arxiv.org/abs/2411.02432</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=42899486</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42899486</guid></item><item><title><![CDATA[New comment by ucha in "'Obelisks': New class of life has been found in human digestive system"]]></title><description><![CDATA[
<p>Does this support the Selfish Gene theory of Richard Dawkins? They look like the smallest self-replicating molecules that he mentions in the beginning of the book.</p>
]]></description><pubDate>Mon, 30 Dec 2024 22:10:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=42554093</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=42554093</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42554093</guid></item><item><title><![CDATA[New comment by ucha in "Is Iceland getting ready to join the EU?"]]></title><description><![CDATA[
<p>Maybe you're polling the people that wanted to join before 2022? Otherwise, how would the jump in support in 2022 be explained?</p>
]]></description><pubDate>Sun, 29 Dec 2024 22:08:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=42543800</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=42543800</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42543800</guid></item><item><title><![CDATA[US man jailed for cloning giant sheep for trophy hunting]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.bbc.com/news/articles/cdenrg7wy5ro">https://www.bbc.com/news/articles/cdenrg7wy5ro</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=41729252">https://news.ycombinator.com/item?id=41729252</a></p>
<p>Points: 36</p>
<p># Comments: 30</p>
]]></description><pubDate>Thu, 03 Oct 2024 10:31:50 +0000</pubDate><link>https://www.bbc.com/news/articles/cdenrg7wy5ro</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=41729252</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41729252</guid></item><item><title><![CDATA[New comment by ucha in "Qualcomm Approached Intel About a Takeover in Recent Days"]]></title><description><![CDATA[
<p>Koyfin and yCharts do that. You're looking for "total return".</p>
]]></description><pubDate>Fri, 20 Sep 2024 21:47:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=41605790</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=41605790</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41605790</guid></item><item><title><![CDATA[Mario Draghi outlines his plan to make Europe more competitive]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.economist.com/by-invitation/2024/09/09/mario-draghi-outlines-his-plan-to-make-europe-more-competitive">https://www.economist.com/by-invitation/2024/09/09/mario-draghi-outlines-his-plan-to-make-europe-more-competitive</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=41500887">https://news.ycombinator.com/item?id=41500887</a></p>
<p>Points: 3</p>
<p># Comments: 1</p>
]]></description><pubDate>Tue, 10 Sep 2024 14:04:22 +0000</pubDate><link>https://www.economist.com/by-invitation/2024/09/09/mario-draghi-outlines-his-plan-to-make-europe-more-competitive</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=41500887</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41500887</guid></item><item><title><![CDATA[What Physicists Have Been Missing]]></title><description><![CDATA[
<p>Article URL: <a href="https://nautil.us/what-physicists-have-been-missing-506607/">https://nautil.us/what-physicists-have-been-missing-506607/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=39691046">https://news.ycombinator.com/item?id=39691046</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 13 Mar 2024 13:31:32 +0000</pubDate><link>https://nautil.us/what-physicists-have-been-missing-506607/</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=39691046</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39691046</guid></item><item><title><![CDATA[How Does the Largest Hedge Fund Make Its Money?]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.nytimes.com/2023/11/01/business/how-does-the-worlds-largest-hedge-fund-really-make-its-money.html">https://www.nytimes.com/2023/11/01/business/how-does-the-worlds-largest-hedge-fund-really-make-its-money.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=38111477">https://news.ycombinator.com/item?id=38111477</a></p>
<p>Points: 3</p>
<p># Comments: 1</p>
]]></description><pubDate>Thu, 02 Nov 2023 10:33:56 +0000</pubDate><link>https://www.nytimes.com/2023/11/01/business/how-does-the-worlds-largest-hedge-fund-really-make-its-money.html</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=38111477</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38111477</guid></item><item><title><![CDATA[Atomic Gardening]]></title><description><![CDATA[
<p>Article URL: <a href="https://en.wikipedia.org/wiki/Atomic_gardening">https://en.wikipedia.org/wiki/Atomic_gardening</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=37091877">https://news.ycombinator.com/item?id=37091877</a></p>
<p>Points: 18</p>
<p># Comments: 1</p>
]]></description><pubDate>Fri, 11 Aug 2023 18:09:27 +0000</pubDate><link>https://en.wikipedia.org/wiki/Atomic_gardening</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=37091877</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=37091877</guid></item><item><title><![CDATA[New comment by ucha in "Claude 2"]]></title><description><![CDATA[
<p>Wow this sounds like they trained it on ChatGPT outputs.<p>However I could not reproduce this output.</p>
]]></description><pubDate>Tue, 11 Jul 2023 20:11:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=36686601</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=36686601</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=36686601</guid></item><item><title><![CDATA[New comment by ucha in "US 5-year credit default swaps rise to highest since 2011"]]></title><description><![CDATA[
<p>Relevant Matt Levine: "Why Would Anyone Buy Credit Default Swaps on the U.S.?"<p><a href="https://archive.is/r0lI2" rel="nofollow">https://archive.is/r0lI2</a></p>
]]></description><pubDate>Fri, 21 Apr 2023 23:14:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=35661504</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=35661504</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35661504</guid></item><item><title><![CDATA[New comment by ucha in "The U.S. cracked a $3.4B crypto heist and Bitcoin’s anonymity"]]></title><description><![CDATA[
<p>Thanks for explaining this!<p>It seemed to me that option #1 only applied if you were a fugitive and your whereabouts were unknown, not when you simply leave the state or country.<p>As to furtherance, I wonder how precise the definition of it is; is using proceeds of a crime, furtherance of a crime? Is obfuscating a crime, furtherance of a crime?<p>Edit: This answers most of my questions and the answer seems to be "it's complicated"
<a href="https://sgp.fas.org/crs/misc/RL31253.pdf" rel="nofollow">https://sgp.fas.org/crs/misc/RL31253.pdf</a></p>
]]></description><pubDate>Wed, 12 Apr 2023 23:48:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=35548759</link><dc:creator>ucha</dc:creator><comments>https://news.ycombinator.com/item?id=35548759</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35548759</guid></item></channel></rss>