<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: kurthr</title><link>https://news.ycombinator.com/user?id=kurthr</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 21 Nov 2026 01:49:31 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=kurthr" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by kurthr in "Synology Security Advisory"]]></title><description><![CDATA[
<p>LLMs hits all of the top 3!<p>I mean, I'm kinda kidding, but #1 is #1.</p>
]]></description><pubDate>Fri, 20 Nov 2026 01:52:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=50569172</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50569172</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50569172</guid></item><item><title><![CDATA[New comment by kurthr in "Jensen Huang might return to Nvidia as CEO"]]></title><description><![CDATA[
<p>Everyone in a position to do something about this has known for more than a year. Him, his wife, the board, the FBI, Commerce, Treasury, the president, the media.<p>No it wasn't reported by the NYT that he had an affair until this month, but plenty of people have known, talked publicly about, and ignored his relationships AND the chip sales for months and months, until... what? Someone decided it was time for a tabloid storm.<p><a href="https://daringfireball.net/linked/2026/11/09/culper-research-huang-huang-report" rel="nofollow">https://daringfireball.net/linked/2026/11/09/culper-research...</a><p>My pet theory is that Jensen & Wife want to sell some shares and this is useful. Alternative is a government or media operative, who thinks there's hay to be made.</p>
]]></description><pubDate>Mon, 16 Nov 2026 00:06:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=50512323</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50512323</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50512323</guid></item><item><title><![CDATA[New comment by kurthr in "Your TV can see through walls (2025)"]]></title><description><![CDATA[
<p>Just toss a doppler laser interferometer on a window, mirror, etc. At least with fiber-optics you need physical access. Single mode (I think that's most of FTTH) is pretty insensitive to vibration and hard to localize so you'd really have to install something.</p>
]]></description><pubDate>Tue, 10 Nov 2026 08:38:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=50437612</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50437612</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50437612</guid></item><item><title><![CDATA[New comment by kurthr in "ChatGPT Keeps Telling People 404 Media Can Solve Their Tech Support Problems"]]></title><description><![CDATA[
<p>Just talking to some InfoSec guys at a party on Saturday at noon, they couldn't drink because they were headed back into work. It's not just PHP, it's everything. They were "joking" about changing professions, because it sucks and it's getting worse. They can't even use frontier models, just use open weights to not get degraded.</p>
]]></description><pubDate>Mon, 26 Oct 2026 19:59:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=50250040</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50250040</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50250040</guid></item><item><title><![CDATA[New comment by kurthr in "Anthropic S-1"]]></title><description><![CDATA[
<p>Is it the EGC lack of disclosure rules (for pre-Unicorns), the fractional share ownership? How is the JOBS Act what drove change?</p>
]]></description><pubDate>Mon, 26 Oct 2026 19:26:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=50249565</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50249565</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50249565</guid></item><item><title><![CDATA[New comment by kurthr in "A mathematician reads the AI future"]]></title><description><![CDATA[
<p>There's lots of stories of non-anthropomorphic intelligences that are human friendly. Sometimes humans do bad things with them, but they're not evil. "Shall we play a game?", HAL9000, Phaecian ships of antiquity, or those of the Culture. Androids seem uniquely unfortunate, perhaps as another poster noted, because of the stories of golem.</p>
]]></description><pubDate>Mon, 26 Oct 2026 14:21:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=50244201</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50244201</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50244201</guid></item><item><title><![CDATA[New comment by kurthr in "A mathematician reads the AI future"]]></title><description><![CDATA[
<p>Well, yes stories is basically what LLMs are, and they're optimized to be stories that humans like. I fully expect to be entertained to death before the droid wars, but if there is a horribly misaligned LLM, I fully expect it to be created by the guy who thinks bipedal anthropomorphic robots are a good idea.</p>
]]></description><pubDate>Mon, 26 Oct 2026 14:08:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=50243985</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50243985</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50243985</guid></item><item><title><![CDATA[New comment by kurthr in "Anthropic S-1"]]></title><description><![CDATA[
<p>Wait, it fell from 140% growth to 45% growth and that's a good thing? If they fall to 15% growth in Q4, will that be even better? Where are we supposed to be on the adoption curve?</p>
]]></description><pubDate>Mon, 26 Oct 2026 13:34:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=50243423</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50243423</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50243423</guid></item><item><title><![CDATA[New comment by kurthr in "Birdwatching rewires the brain and slows cognitive decline"]]></title><description><![CDATA[
<p>What percentage of people are neither of the simple common XY nor XX chromosomes? 
It's about as many people as there are with red hair. If you don't include chimerics it's about 0.25-1.7%, if you do include twin chimerics it's 8%, not including mothers micro-chimerism.<p>Probably shouldn't be having those Kenyans (Kalenjin) competing in marathons either. Completely unfair advantage.<p>All models are wrong, some are useful.</p>
]]></description><pubDate>Mon, 26 Oct 2026 13:24:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=50243233</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50243233</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50243233</guid></item><item><title><![CDATA[New comment by kurthr in "Scientists rediscover how to make ancient Roman colour-changing glass"]]></title><description><![CDATA[
<p>Note that rubies are also color changing (thermochromatic)! The absorption band gap of ruby (well, really the Cr d-shell band) expands to longer wave lengths at high temp and contracts at low temperatures so that it is light orange ~77K or -200C and turns black (IR) ~400C.</p>
]]></description><pubDate>Mon, 26 Oct 2026 13:04:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=50242926</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50242926</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50242926</guid></item><item><title><![CDATA[New comment by kurthr in "I don't understand quantum physics (2018)"]]></title><description><![CDATA[
<p>Yeah, I don't understand where this is coming from. If you're not teaching the math, what are you even doing? Maybe it's from pop-sci books or something where they're really selling woo rather than comprehension?<p>My books went from 2 slit experiments, to particle in a box (1d and then 2d harmonic oscillators) with tunneling, and then the Hydrogen atom potential with electron orbitals as an extension, to perturbations for He, to Hartree-Fock and Born-Oppenheimer for large atoms and molecular models. Learning the <bra/ket> formalism can be useful for thinking about it, but it has to be motivated by the math.<p>I guess black body and scattering were pretty early too as motivation for the developments (as they historically were). Whereas EPR/Bell, Decoherence, and "Interpretations" were mostly discussed as interesting flavor and philosophical arguments.</p>
]]></description><pubDate>Sun, 25 Oct 2026 19:32:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=50235237</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50235237</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50235237</guid></item><item><title><![CDATA[New comment by kurthr in "A mathematician reads the AI future"]]></title><description><![CDATA[
<p>It's what plants crave!<p>To add a bit more almost off topic substance from the enshitification of human work. I really do think the idea of hooking these "AI" models up to humanoid robots is a bad idea not for all the mundane reasons, but just because "alignment" seems impossible. All of the world's stories about humanoid robots basically involve them attacking humans and behaving in nefarious destructive ways. It's fundamentally embedded in the models, and there's no reason post training would pull it out, because it's basically in the key value vector meaning "malicious human destroyer". Even to find counter examples like C3PO, Daneel Olivaw, Data, or Wall-ee they come after droid/robot driven wars and apocolypses. It seems unwise.</p>
]]></description><pubDate>Sun, 25 Oct 2026 17:21:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=50233842</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50233842</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50233842</guid></item><item><title><![CDATA[New comment by kurthr in "Imagine an AI future where being banned means you are now economically useless"]]></title><description><![CDATA[
<p>They'll turn off your lifesaving drone internet access as well.</p>
]]></description><pubDate>Sun, 25 Oct 2026 17:19:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=50233824</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50233824</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50233824</guid></item><item><title><![CDATA[New comment by kurthr in "Physicists trap a single antimatter particle for a month"]]></title><description><![CDATA[
<p>And yet the lifetime of free "thermal" neutrons is unknown (there are two measurements that give results off by ~1%) between 888 and 878 seconds.<p>Notably, neutrons are quite unstable outside of a nucleus and decay generating a proton, electron, and a neutrino. There are many theories on the discrepancy.<p><a href="https://journals.aps.org/prd/abstract/10.1103/PhysRevD.110.073004" rel="nofollow">https://journals.aps.org/prd/abstract/10.1103/PhysRevD.110.0...</a></p>
]]></description><pubDate>Sun, 25 Oct 2026 05:23:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=50228581</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50228581</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50228581</guid></item><item><title><![CDATA[New comment by kurthr in "AI analysis of CT scans finds 42.9% of patients have coronary inflammation"]]></title><description><![CDATA[
<p>Actually, yes guidelines are actually that any prior statin prescription removes the need for CAC measurements. That includes those who are discovered to have statin intolerance, and thus requires out-of-pocket payment for any CAC.<p>I think the challenge here is that other than a CCTA there's no good way to estimate individual (as opposed to hereditary) ASCVD risk. Sure LDL, ApoB, Lp(a), and BP, are statistical proxies, but tell you little about the individual risk. CAC isn't much better, but if you've been taking EPA and VitK2 you have a much higher chance of more stable plaques, which are also more visible on CAC without much larger X-ray exposure and iodine contrast effects.</p>
]]></description><pubDate>Thu, 22 Oct 2026 20:39:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=50200708</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50200708</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50200708</guid></item><item><title><![CDATA[New comment by kurthr in "Oregon lawmakers approve money to buy an iconic waterfall listed on Redfin"]]></title><description><![CDATA[
<p>You can't optimize the sweet sweet ad cash by just being a journalist! The need for a commercial tie-in zinger twist is ingrained now.</p>
]]></description><pubDate>Wed, 21 Oct 2026 20:33:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50187221</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50187221</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50187221</guid></item><item><title><![CDATA[New comment by kurthr in "DeepInfer: High-performance inference framework for DeepSeek models"]]></title><description><![CDATA[
<p>To steel man the argument, I think that there really is a Moore's Cliff coming. I don't think that resolution is going to matter as much, although process/design knowledge will still be a huge barrier. I don't see even 10A being as big a node as 14A or 2nm. Too many things fall apart all at once, and the cyclical nature of semi will raise it's bloody head above the industry as the AI trade bursts. So many $10Bs will be spent and margins will plummet... again, just in time for China to catch up.<p>Already, the big capabilities are becoming multi-layer/die integration. Whether it's BSPDN, stacked die, optical interconnects, etc it won't be as simple as Denard scaling. It will be dependent on experience and supply chains, which are harder to transfer than equipment or IP.<p>I do think the market for inference will drive a lot of innovation, and a lot of that will be at larger nodes. A single reticle die masked ROM with CIM can do 7B at 6nm. At 2-3nm you can do 27-32B which is a single expert or router. At that point there's not much reason to shrink for AI since it's mostly 1-hot die per token (except for the router). Update the parameters every 3-6mo when a new frontier comes out, and take the cost/power savings to the bank.</p>
]]></description><pubDate>Wed, 21 Oct 2026 19:18:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=50186099</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50186099</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50186099</guid></item><item><title><![CDATA[New comment by kurthr in "DeepInfer: High-performance inference framework for DeepSeek models"]]></title><description><![CDATA[
<p>Go bet on it @=^D It's not speculation, it's what they sell to the world and build processors for globally available Huawei/Honor phones. It's not a secret.<p>If they had a solution to get to >90% yields on sub 5nm they'd be doing it now.
They don't, even the 7-10nm isn't there. Best hope is they get to buy the entire world's supply of EUV in 2027, then they install all of them, then they bring them up, then they build processes around them and make masks, then they ramp the yield on 3-4nm by copying exact TSMC.<p>The hilarious thing is that TSMC couldn't do this at a new fab in Taiwan, with full access to equipment, supply chain, and people. So, nope.</p>
]]></description><pubDate>Wed, 21 Oct 2026 02:08:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=50175212</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50175212</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50175212</guid></item><item><title><![CDATA[New comment by kurthr in "DeepInfer: High-performance inference framework for DeepSeek models"]]></title><description><![CDATA[
<p>Maybe they are assembled in China?<p>You're not going to make money buying a new 14nm bitcoin miner, and that is where the China foundries are for cost effective silicon. They're great at 28nm+ and good at 22nm pricing.<p>Most new design modern sub 7nm ASICs are made in Taiwan/Korea by TSMC and SLSI. That is where miners are profitable.</p>
]]></description><pubDate>Tue, 20 Oct 2026 17:00:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=50169348</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50169348</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50169348</guid></item><item><title><![CDATA[New comment by kurthr in "Americans Are Spending More Money to Drive Worse Cars"]]></title><description><![CDATA[
<p>Just because a group claims an -ist or puts an -ism in their name doesn't make them so. The NAZIs weren't socialist, they locked them up. The CCP isn't communist except in name. Russia has been an oligarchy since before the wall fell.<p>And what we have in the US right now is not a free market, nor a well regulated and efficient one. You can call it crony capitalism, if you want, but when banks are bailed out to the tune of $Ts, Federal deficit spending is responsible for "positive" economic growth, and speculative companies are shoved into the indexes that make up what savings the working class has, it barely makes the cut.<p>Plenty of more "socialist" places with with free healthcare and rule of law, are more market capitalist. Enshitification happens where it is allowed to, because it is allowed to.</p>
]]></description><pubDate>Tue, 20 Oct 2026 05:54:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=50161874</link><dc:creator>kurthr</dc:creator><comments>https://news.ycombinator.com/item?id=50161874</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50161874</guid></item></channel></rss>