<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: kjshsh123</title><link>https://news.ycombinator.com/user?id=kjshsh123</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 04 Sep 2026 09:22:20 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=kjshsh123" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by kjshsh123 in "How concerned should we be about Astra's recurrent architecture?"]]></title><description><![CDATA[
<p>On further thought, I think the author's intent was to say that classic RNNs have "unbounded temporal accumulation in the hidden state".</p>
]]></description><pubDate>Thu, 03 Sep 2026 19:54:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=49555886</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=49555886</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49555886</guid></item><item><title><![CDATA[New comment by kjshsh123 in "How concerned should we be about Astra's recurrent architecture?"]]></title><description><![CDATA[
<p>>In contrast to a classic RNN, there's no unbounded hidden state accumulating across an entire trajectory<p>I don' understand this line. In a classic RNN hidden state is <i>bounded</i> dimension. In fact it's transformers that technically have unbounded hidden state.<p>You can't parallelize classic nonlinear RNNs for various reasons but in training both RNN and Transformer depend on the entire sequence history in a way that is unbounded. Of course in practice you just train on a max sequence length.<p>RNN xhat[t+1]=f(x[t],h[t])<p>Transformer/self-attention xhat[t+1]=f(x[t],h[t],h[t-1],...,h[1])</p>
]]></description><pubDate>Thu, 03 Sep 2026 19:49:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=49555757</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=49555757</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49555757</guid></item><item><title><![CDATA[New comment by kjshsh123 in "You probably own this 7-Eleven (and that's why it looks so sad)"]]></title><description><![CDATA[
<p>Land value tax would solve this.</p>
]]></description><pubDate>Mon, 31 Aug 2026 19:53:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=49514176</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=49514176</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49514176</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Claude users are mad that Anthropic's new watermarks will catch them using it"]]></title><description><![CDATA[
<p>Could probably make a 1B model that removes watermarks. Chooses synonyms, trains on user style, changes order around etc.</p>
]]></description><pubDate>Thu, 13 Aug 2026 12:25:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=49284897</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=49284897</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49284897</guid></item><item><title><![CDATA[Supply and Demand Is Not What Most People Think]]></title><description><![CDATA[
<p>Article URL: <a href="https://shonczinner.substack.com/p/supply-and-demand-is-not-what-most">https://shonczinner.substack.com/p/supply-and-demand-is-not-what-most</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49134511">https://news.ycombinator.com/item?id=49134511</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 01 Aug 2026 13:58:21 +0000</pubDate><link>https://shonczinner.substack.com/p/supply-and-demand-is-not-what-most</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=49134511</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49134511</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Claude Fable produced a counterexample to the Jacobian Conjecture"]]></title><description><![CDATA[
<p>He's not dead, but it is bad to gossip regardless.</p>
]]></description><pubDate>Mon, 20 Jul 2026 13:25:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=48978563</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48978563</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48978563</guid></item><item><title><![CDATA[Our Straussian Technocracy]]></title><description><![CDATA[
<p>Article URL: <a href="https://libertiesjournal.com/online-articles/our-straussian-techocracy/">https://libertiesjournal.com/online-articles/our-straussian-techocracy/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48850926">https://news.ycombinator.com/item?id=48850926</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 09 Jul 2026 19:08:26 +0000</pubDate><link>https://libertiesjournal.com/online-articles/our-straussian-techocracy/</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48850926</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48850926</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Bring back crappy forums"]]></title><description><![CDATA[
<p>Like I said, these are negative internalities. I don't think most users care about these negative impacts on themselves.</p>
]]></description><pubDate>Thu, 02 Jul 2026 13:27:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=48761271</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48761271</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48761271</guid></item><item><title><![CDATA[New comment by kjshsh123 in "We Don't Have to Be This Bad at Improving Society"]]></title><description><![CDATA[
<p>Most countries don't allow billionaires to influence elections like the USA. Maybe they still do in other ways.<p>Anyway, I think blaming the 0.01% for how 50%+1 vote is too easy.<p>Take electoral reform. Is that the 0.01%, or is the case that it's just the only one that could improve elections with electoral reform never has any incentive? Still, if voters really made it their priority and punished parties that run on it and ignore it once they win, it would be more likely to happen.<p>No doubt there is problems of reproduction of class, and conservatism to preserve existing hierarchies as they are. But is it the 0.01%? Or is it actually a much larger fraction? Tyranny of the majority is still tyranny even if it's 51% lording it over the 49% instead of 0.01% lording it over the 99.99%.<p>History shows people have a desire for minority scapegoats. We should know that is wrong.<p>Take housing. Is unaffordability to blame on the 5% of the population that are landlords? The 1% involved in housing development? Or is it maybe the 60%+ households that live in detached homes being insulated from the problem and being NIMBYs in local elections?</p>
]]></description><pubDate>Thu, 02 Jul 2026 13:25:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=48761228</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48761228</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48761228</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Bring back crappy forums"]]></title><description><![CDATA[
<p>Social media won because it's better for the consumer and producer.<p>For the producer, it's free infrastructure but it's also advertising. Having a large subreddit means your game getting recommended to others and potentially being seen being introduced to more people.<p>For the consumer, these social media sites do usually do provide a better experience in showing people what they want to see and keeping away stuff they don't.<p>I'm sympathetic to forums just because I think if someone likes something they shouldn't need to join a potentially social media site with potentially toxic designs and sub-communities. But these are negative internalities that people mostly ignore.</p>
]]></description><pubDate>Thu, 02 Jul 2026 04:07:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=48756445</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48756445</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48756445</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Why I Stopped Arguing with People"]]></title><description><![CDATA[
<p>>If you genuinely believe something others don’t, that’s not a debate to win. That’s an edge. The market rewards being right in a way that no argument ever will. Instead of persuading the skeptic, ship the thing they think is wrong and let reality settle it.<p>I wonder if victims of religious persecution agree...<p>Facts do not always win. The evolutionary fitness of an idea is (sadly) not entirely dictated by its truthfulness.</p>
]]></description><pubDate>Wed, 01 Jul 2026 16:08:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=48749150</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48749150</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48749150</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Age verification is just a precursor to automated attribution of speech"]]></title><description><![CDATA[
<p>That sounds like mistaken optimism due to a mistaken interpretation of the invisible hand.<p>Mandating age verification and the inevitable implementation requirements are bad for freedom.<p>Behaviour changes and innovations will mitigate some of the negatives, but bad things are bad.</p>
]]></description><pubDate>Mon, 29 Jun 2026 04:53:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=48714952</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48714952</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48714952</guid></item><item><title><![CDATA[Neural Cellular Automata and Recurrent Architectures]]></title><description><![CDATA[
<p>Article URL: <a href="https://shonczinner.github.io/posts/gameoflife/">https://shonczinner.github.io/posts/gameoflife/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48611548">https://news.ycombinator.com/item?id=48611548</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 20 Jun 2026 18:21:48 +0000</pubDate><link>https://shonczinner.github.io/posts/gameoflife/</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48611548</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48611548</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Ozempic and Wegovy linked to surprising drop in violent behavior"]]></title><description><![CDATA[
<p>It's only a matter of time before nootropics forums start experimenting with these drugs...<p>These drugs seem to change something fundamental that reduces problematic behavior.</p>
]]></description><pubDate>Thu, 18 Jun 2026 17:53:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=48588966</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48588966</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48588966</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Apple boss Tim Cook says prices to rise due to memory chip costs"]]></title><description><![CDATA[
<p>>The Alchian–Allen effect was described in 1964 by Armen Alchian and William R Allen in the book University Economics (now called Exchange and Production[1]). It states that when the prices of two substitute goods, such as high and low grades of the same product, are both increased by a fixed per-unit amount such as a transportation cost or a lump-sum tax, consumption will shift toward the higher-grade product. This is because the added per-unit amount decreases the relative price of the higher-grade product.<p><a href="https://en.wikipedia.org/wiki/Alchian%E2%80%93Allen_effect?wprov=sfla1" rel="nofollow">https://en.wikipedia.org/wiki/Alchian%E2%80%93Allen_effect?w...</a></p>
]]></description><pubDate>Thu, 18 Jun 2026 14:39:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=48586122</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48586122</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48586122</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Peopleless economy? Not technically impossible"]]></title><description><![CDATA[
<p><a href="https://en.wikipedia.org/wiki/Consumption_%28economics%29?wprov=sfla1" rel="nofollow">https://en.wikipedia.org/wiki/Consumption_%28economics%29?wp...</a><p>>Thus no matter what your productive output is per hour of labour, you will always work as much as you can, because you are - presumably - insatiably driven to always consume ever more with no end to it.<p>No, that's not how marginalism works. There's both diminishing marginal utility from consumption, and leisure.</p>
]]></description><pubDate>Wed, 17 Jun 2026 23:25:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=48578441</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48578441</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48578441</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Peopleless economy? Not technically impossible"]]></title><description><![CDATA[
<p>The person I replied to was trying to separate necessities from consumption in the "traditional sense of the word".<p>No doubt there's problematic regulations like exclusionary zoning laws. But you can't say that these regulations are so binding that there is no choice and no expression of preferences as opposed to needs in their choices. Lots of homes still have unmandated second floors, basements, bathrooms, and square footage.<p>The human brain is great at (ir)rationalizing wants as needs. If you want to live in a nice place in a high-cost of living city, that's a want, not a need.</p>
]]></description><pubDate>Tue, 16 Jun 2026 13:34:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=48555083</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48555083</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48555083</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Labor Is a Market Distortion, we need VAT and UBI"]]></title><description><![CDATA[
<p>That's not a distortion, that's just what the choice is.<p>When a new gaming console comes out and someone works extra shifts to afford it sooner, that's not a distortion.<p>I think your mistake, based on how you're talking about the "real value of leisure" is thinking there is a single value of x. That's labor theory of value era thinking. There is only marginal utility and marginal rates of substitution.<p>Trading leisure for consumption via labor is the choice, not a distortion.</p>
]]></description><pubDate>Tue, 16 Jun 2026 13:04:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=48554675</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48554675</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48554675</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Peopleless economy? Not technically impossible"]]></title><description><![CDATA[
<p>Necessities count as consumption. You could survive off rice and live in an internet cafe for $15/night. I don't know what you think the traditional sense of the word "consumption" is.</p>
]]></description><pubDate>Tue, 16 Jun 2026 12:11:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=48553995</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48553995</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48553995</guid></item><item><title><![CDATA[New comment by kjshsh123 in "Labor Is a Market Distortion, we need VAT and UBI"]]></title><description><![CDATA[
<p>I don't really. It smells to me like the lump of labor fallacy. Labor participation results from supply and demand and it going down does not necessarily indicate less demand, but can also indicate less supply. That's what you'd expect as countries and the world become wealthier. Leisure is a normal good.</p>
]]></description><pubDate>Sat, 13 Jun 2026 14:16:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=48517603</link><dc:creator>kjshsh123</dc:creator><comments>https://news.ycombinator.com/item?id=48517603</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48517603</guid></item></channel></rss>