<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: sanxiyn</title><link>https://news.ycombinator.com/user?id=sanxiyn</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 20 Nov 2026 18:47:32 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=sanxiyn" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by sanxiyn in "Show HN: Fursoy – Out-of-order RV64IMAFDV core in SystemVerilog (Linux 6.6)"]]></title><description><![CDATA[
<p>> for verification there's a reference model written in Sail you can compare against (though tbf it sounds like he didn't)<p>Can you elaborate on "it sounds like he didn't" because he clearly did? There is a Sail testbench at verification/s1_sail_cosim_tb.cpp.</p>
]]></description><pubDate>Sun, 15 Nov 2026 01:08:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50502910</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=50502910</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50502910</guid></item><item><title><![CDATA[New comment by sanxiyn in "What should we tell our students?"]]></title><description><![CDATA[
<p>I am not sure what you are trying to imply, but quasi-Riemann hypothesis is in fact not Riemann hypothesis and resolution of the former, while truly fucking amazing (it wouldn't be an exaggeration to say it is a discovery of century), gives no hint how to resolve the later.</p>
]]></description><pubDate>Fri, 09 Oct 2026 04:21:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=50015960</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=50015960</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50015960</guid></item><item><title><![CDATA[New comment by sanxiyn in "OpenAI, the Partition Principle, and Mathematics"]]></title><description><![CDATA[
<p>As I said, a proper eval is better, but it measures something real that gives a good training signal, and there is lack of good alternatives for math eval. Since the result is 372/4000, it is also unsaturated.</p>
]]></description><pubDate>Fri, 09 Oct 2026 01:32:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50014885</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=50014885</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50014885</guid></item><item><title><![CDATA[New comment by sanxiyn in "OpenAI, the Partition Principle, and Mathematics"]]></title><description><![CDATA[
<p>OpenAI in fact didn't know what to do with results and didn't want to flood the world, so they asked mathematicians. Mathematicians recommended OpenAI to release them. My guess is it would have been better for OpenAI if they didn't release them. OpenAI is basically doing this as a goodwill.<p>(See <a href="https://agmai.org/general-sep29/" rel="nofollow">https://agmai.org/general-sep29/</a> for the recommendation in question.)<p>People seem to have very misguided ideas about why OpenAI is doing this at all. It is not to brag or to torture mathematicians. It is an eval. OpenAI is known to be willing to pay large amount of money to get a good eval, think FrontierMath. FrontierMath is now saturated, so they need a replacement eval for math. Open math problems are actually a fairly good eval, although a proper eval is better (eg FrontierMath has known difficulty and have tiers from 1 to 4).<p>Mathematicians would prefer if OpenAI didn't use open math problems as an eval, but OpenAI is not obliged. I actually think OpenAI wouldn't point AI to open math problems if unsaturated FrontierMath Super Duper is available, as it just angers mathematicians, but such eval is not in fact available. Given OpenAI used open math problems as an eval, they could just throw out the result (this is in fact better as an eval since it will keep problems useful longer), but mathematicians preferred to see the result. So OpenAI released them.</p>
]]></description><pubDate>Fri, 09 Oct 2026 00:58:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=50014661</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=50014661</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50014661</guid></item><item><title><![CDATA[New comment by sanxiyn in "OpenAI withdraws three mathematical results"]]></title><description><![CDATA[
<p>Yes: <a href="https://x.com/ElliotGlazer/status/2108026240582246600" rel="nofollow">https://x.com/ElliotGlazer/status/2108026240582246600</a></p>
]]></description><pubDate>Thu, 08 Oct 2026 11:58:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=50004774</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=50004774</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50004774</guid></item><item><title><![CDATA[New comment by sanxiyn in "“Math 2.0” will need to value mathematical progress more holistically"]]></title><description><![CDATA[
<p>This is basically On proof and progress in mathematics by Thurston restated. When Thurston wrote it in 1994, many people didn't understand what he is talking about.<p><a href="https://arxiv.org/abs/math/9404236" rel="nofollow">https://arxiv.org/abs/math/9404236</a></p>
]]></description><pubDate>Thu, 08 Oct 2026 06:25:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=50002418</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=50002418</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50002418</guid></item><item><title><![CDATA[New comment by sanxiyn in "Early rogue AI agent activity and attempts to hack found on urlquery.net"]]></title><description><![CDATA[
<p>This in fact already happened (exactly OpenClaw, even).<p>AI assistant hacks gym website in first known Australian autonomous cyber attack: <a href="https://www.abc.net.au/news/2026-08-10/ai-assistant-hacks-gym-website-aus-cyber-attack/107007986" rel="nofollow">https://www.abc.net.au/news/2026-08-10/ai-assistant-hacks-gy...</a><p>General opinion at the time was it was in fact ambiguous who was legally liable.</p>
]]></description><pubDate>Thu, 24 Sep 2026 08:18:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=49827765</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49827765</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49827765</guid></item><item><title><![CDATA[New comment by sanxiyn in "OpenAI breaches Medicare, Albanese reveals"]]></title><description><![CDATA[
<p>For technical details, see <a href="https://transluce.org/agent-activity" rel="nofollow">https://transluce.org/agent-activity</a>.</p>
]]></description><pubDate>Thu, 24 Sep 2026 05:48:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=49826700</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49826700</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49826700</guid></item><item><title><![CDATA[AI Cheating Is on the Rise]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.vals.ai/blogs/cheating-on-the-rise">https://www.vals.ai/blogs/cheating-on-the-rise</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49735030">https://news.ycombinator.com/item?id=49735030</a></p>
<p>Points: 4</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 17 Sep 2026 00:42:25 +0000</pubDate><link>https://www.vals.ai/blogs/cheating-on-the-rise</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49735030</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49735030</guid></item><item><title><![CDATA[New comment by sanxiyn in "Artificial Analysis Intelligence Index v4.2"]]></title><description><![CDATA[
<p>It is very unfortunate they upweighted SciCode from 8% to 10%. SciCode is a broken benchmark: see <a href="https://arxiv.org/abs/2608.04975" rel="nofollow">https://arxiv.org/abs/2608.04975</a>.</p>
]]></description><pubDate>Sat, 05 Sep 2026 05:40:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=49573404</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49573404</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49573404</guid></item><item><title><![CDATA[New comment by sanxiyn in "Formalizing Fermat's Last Theorem"]]></title><description><![CDATA[
<p>Lean's three standard axioms are documented in The Lean Language Reference.<p><a href="https://lean-lang.org/doc/reference/latest/Axioms/#standard-axioms" rel="nofollow">https://lean-lang.org/doc/reference/latest/Axioms/#standard-...</a><p>The axiom of choice: axiom Classical.choice {α : Sort u} : Nonempty α → α<p>The axiom of propositional extensionality: axiom propext {a b : Prop} : (a ↔ b) → a = b<p>The quotient axiom: axiom Quot.sound : ∀ {α : Sort u} {r : α → α → Prop} {a b : α}, r a b → Eq (Quot.mk r a) (Quot.mk r b)</p>
]]></description><pubDate>Fri, 04 Sep 2026 23:40:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=49571477</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49571477</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49571477</guid></item><item><title><![CDATA[New comment by sanxiyn in "Formalizing Fermat's Last Theorem"]]></title><description><![CDATA[
<p>New proof: The Classification of the Finite Simple Groups (American Mathematical Society Mathematical Surveys and Monographs vol. 40).<p><a href="https://www.ams.org/publications/authors/books/postpub/surv-40" rel="nofollow">https://www.ams.org/publications/authors/books/postpub/surv-...</a><p>Number 1 (1994), Number 2 (1995), Number 3 (1997), Number 4 (1999), Number 5 (2002), Number 6 (2004), Number 7 (2018), Number 8 (2018), Number 9 (2021), Number 10 (2023). 10 volumes and >4000 pages so far, number 11 is in progress, and end is in sight, probably two more volumes or so.<p><a href="https://www.ams.org/journals/notices/201806/rnoti-p646.pdf" rel="nofollow">https://www.ams.org/journals/notices/201806/rnoti-p646.pdf</a><p>People were curious what is going on during 2004-2018. A progress report was published in 2018 right before publication of number 7 and 8. In a sense it was the peak, number 8 completes the proof of so-called "generic case". The rest is "special case". It doesn't mean things get easier, but in some specific sense number 8 completed proof for almost all groups.<p>Now new proof's end is in sight, people are planning new new proof.</p>
]]></description><pubDate>Fri, 04 Sep 2026 23:15:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=49571295</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49571295</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49571295</guid></item><item><title><![CDATA[New comment by sanxiyn in "Formalizing Fermat's Last Theorem"]]></title><description><![CDATA[
<p>Yes, but Claude formalized a different proof than Buzzard is trying to, so it helps less than you think. (It certainly helps!)</p>
]]></description><pubDate>Fri, 04 Sep 2026 22:55:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=49571156</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49571156</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49571156</guid></item><item><title><![CDATA[New comment by sanxiyn in "Improving our alignment and security efforts"]]></title><description><![CDATA[
<p>I believe it is referring to <a href="https://www.pacingthefrontier.com/" rel="nofollow">https://www.pacingthefrontier.com/</a>.</p>
]]></description><pubDate>Wed, 02 Sep 2026 00:51:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=49530389</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49530389</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49530389</guid></item><item><title><![CDATA[New comment by sanxiyn in "Mathematics in the age of AI"]]></title><description><![CDATA[
<p>The consensus is that proof is in fact incorrect. People tried really hard (like putting in a year of effort) and most converged to the same place, that proof of 3.12 is incorrect or has a gap. Peter Scholze (who won Fields Medal) and Jakob Stix did a writeup. People seem to think Shinichi Mochizuki correctly reduced ABC conjecture to 3.12, but didn't prove 3.12, and also are doubtful about the whole program because 3.12 doesn't seem any easier than ABC conjecture while complicating everything.<p><a href="https://ncatlab.org/nlab/files/why_abc_is_still_a_conjecture.pdf" rel="nofollow">https://ncatlab.org/nlab/files/why_abc_is_still_a_conjecture...</a></p>
]]></description><pubDate>Thu, 20 Aug 2026 00:37:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=49369047</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49369047</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49369047</guid></item><item><title><![CDATA[New comment by sanxiyn in "GPU Offload in Rust: Portable, Safe, and Fast"]]></title><description><![CDATA[
<p>People go through trouble to write Python-shaped DSL for GPU compute. We will go "why o why?", but apparently such things are necessary to succeed in the market.</p>
]]></description><pubDate>Mon, 17 Aug 2026 22:37:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=49338618</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49338618</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49338618</guid></item><item><title><![CDATA[New comment by sanxiyn in "Stripe will reportedly acquire OpenRouter for $7B+"]]></title><description><![CDATA[
<p>I think volume is itself good and help Stripe negotiate lower rate with banks etc.</p>
]]></description><pubDate>Sun, 16 Aug 2026 23:15:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=49324779</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49324779</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49324779</guid></item><item><title><![CDATA[New comment by sanxiyn in "NetBSD 11.0"]]></title><description><![CDATA[
<p>Congratulations to the first NetBSD release with RISC-V port!</p>
]]></description><pubDate>Sun, 02 Aug 2026 00:31:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=49139949</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49139949</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49139949</guid></item><item><title><![CDATA[New comment by sanxiyn in "Investigating three real-world incidents in our cybersecurity evaluations"]]></title><description><![CDATA[
<p>This is not okay. NSA should audit both OpenAI and Anthropic on national security ground. This seems far more justifiable than Mythos export control.</p>
]]></description><pubDate>Thu, 30 Jul 2026 23:29:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=49117163</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49117163</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49117163</guid></item><item><title><![CDATA[New comment by sanxiyn in "Our position on open-weights models"]]></title><description><![CDATA[
<p>Yes, I agree it is effectively a blanket ban (above some capability) for now. I hope AI alignment research advances in the future so that it is not so.</p>
]]></description><pubDate>Tue, 28 Jul 2026 00:26:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=49077640</link><dc:creator>sanxiyn</dc:creator><comments>https://news.ycombinator.com/item?id=49077640</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49077640</guid></item></channel></rss>