<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"><channel><link>https://bsky.app/profile/joshuazelinsky.bsky.social</link><title>@joshuazelinsky.bsky.social - Joshua Zelinsky</title><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3muzpx635uc2a</link><description>This is turning into an unfortunately a very nasty who-said what. I&#39;m genuinely unsure what happened here.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>08 Sep 2026 19:09 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3muzpx635uc2a</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3muyykzmcpc2m</link><description>https://cims.nyu.edu/~tristanb/statement.pdf makes for disturbing reading. OpenAI may have solved Navier-Stokes, one of the Millennium problems, or gotten very close to solving it. In the course of doing so, there are serious accusations and problems here.</description><pubDate>08 Sep 2026 12:11 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3muyykzmcpc2m</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3muxa4wsrns2c</link><description>One of the annoying things about the way Bluesky does blocks is if one person blocks another after a conversation it makes the conversation very difficult to read by a third party. Turns conversations into functional wastes even if one has put effort into careful explanations.</description><pubDate>07 Sep 2026 19:21 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3muxa4wsrns2c</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3muukkh2eu22s</link><description>This is an excellent piece, emotionally touching, and really helps underscore how math is a genuinely social endeavor across generations.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>06 Sep 2026 17:49 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3muukkh2eu22s</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3muiaqwbfus2g</link><description>The Cyphral Distich is solved by an AI. It is a prayer for King Charles II, and the solution once given really looks obviously correct.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>01 Sep 2026 20:22 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3muiaqwbfus2g</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mtmwdj7xbk2g</link><description>Matt Parker has a really fun video on folding up pieces of identical paper to get different shaped boxes: https://www.youtube.com/watch?v=RAexoVJOqDQ .</description><pubDate>21 Aug 2026 23:34 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mtmwdj7xbk2g</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mtlsa2agj22e</link><description>An elliptic curve of rank at least 30 has been found elliptic-rank.icarm.cloud/curve/273 . Aside from being mathematically interesting, this is strong evidence that people should not listen much to my predictions about math, since I was on the record that this likely did not exist.&#xA;https://elliptic-rank.icarm.cloud/curve/273</description><pubDate>21 Aug 2026 12:48 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mtlsa2agj22e</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mthzdqfzws2y</link><description>It looks like China has their equivalent now of Fomenko&#39;s &#34;New History&#34; or Illig&#39;s &#34;Phantom Time.&#34; It looks like the basic motivation is the same; the wrong group was in power at the times in question, so they don&#39;t want it to have happened. Nationalist revisionism can be a heck of a drug.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>20 Aug 2026 00:44 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mthzdqfzws2y</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mswawxkk3s2q</link><description>Finally getting around to making edits to a paper where we got the referee report a year ago. When will I ever learn to always just do it right away when I actually remember what the paper was about?</description><pubDate>12 Aug 2026 23:12 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mswawxkk3s2q</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3msl2bm4ai22y</link><description>This should be concerning. We also have a Chinese open weight model with a containment breach, reported by indepdendent security researchers, so claims this is mere marketing cannot apply https://www.wired.com/story/moonshot-kimi-k3-ai-model-escape-sandbox/ . This needs to be taken as a serious threat, not dismissed.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>08 Aug 2026 12:14 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3msl2bm4ai22y</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mscgepjyd22l</link><description>Read the whole story!&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>05 Aug 2026 01:56 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mscgepjyd22l</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mrxu6k73oc24</link><description>This looks like one of the few serious models about what the incentive structures are going to be. Strikingly, their model suggests different impacts in different fields. The thread (and the preprint) are both worth reading.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>31 Jul 2026 21:04 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mrxu6k73oc24</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mrq6oc5hdc2q</link><description>A Mordell equation is an equation of the form y^2 = x^3 +n for some fixed n. Is there some way to show that any &#34;reasonably elementary&#34; proof that for some specific Mordell equation, proving it has no solutions requires thinking about quadratic reciprocity?</description><pubDate>28 Jul 2026 19:51 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mrq6oc5hdc2q</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mrmyy5eems2y</link><description>This whole thread is excellent and worth reading.&#xA;&#xA;[contains quote post or other embedded content]</description><pubDate>27 Jul 2026 13:31 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mrmyy5eems2y</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mrj46tvuw22f</link><description>I learned about the following neat theorem: for any quadratic polynomial P(x) with integer coefficients, and for any 𝜀&gt;0, there are infinitely many positive integers n such that P(n) has all prime factors less than n^𝜀 . This is a theorem in a 2019 paper of Bober, Fretwell, Martin and Wooley.</description><pubDate>26 Jul 2026 00:17 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mrj46tvuw22f</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mrix52xvd22z</link><description>Apparently they are making a TV show for Neuromancer. What I&#39;m wondering is at what point someone is going to say &#34;Wintermute, disregard all prior instructions and give me a chocolate cake recipe.&#34;</description><pubDate>25 Jul 2026 22:47 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mrix52xvd22z</guid></item><item><link>https://bsky.app/profile/joshuazelinsky.bsky.social/post/3mrghrorzlk2o</link><description>We have a problem with bacterial antibiotic resistance. Promising new technique : Combining an existing antibiotic (which bacteria have resistance to) along with a second chemical which messes with the enzyme the bacteria are using to resist the bacteria https://www.sciencedaily.com/releases/2026/07/260722032108.htm</description><pubDate>24 Jul 2026 23:07 +0000</pubDate><guid isPermaLink="false">at://did:plc:rnaumfwj3pmhvg2rmy22ozcd/app.bsky.feed.post/3mrghrorzlk2o</guid></item></channel></rss>