<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: beernet</title><link>https://news.ycombinator.com/user?id=beernet</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 20 Nov 2026 08:37:57 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=beernet" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by beernet in "Fable 5.2 one-shotted my nephew's dream game"]]></title><description><![CDATA[
<p>"Look what software/app/... my kid created" appears to become the new "oh look at these cute pictures of my kids". The interest of the forced recipient is equally low, usually.</p>
]]></description><pubDate>Tue, 10 Nov 2026 17:07:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=50443151</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50443151</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50443151</guid></item><item><title><![CDATA[New comment by beernet in "Gemini 4 Carbon is now available"]]></title><description><![CDATA[
<p>For OAI and Anthropic releases we usually get a "pelican" here in the comments pretty (suspiciously) quickly, which then also happens to be on top of the comments. I always found that pretty weird given that's not an actual model "evaluation". It's provocative and gets the people going, I guess.<p>This happens not so much when Google or other labs release (not even on the Argon release post, afaik). Make of that what you will. Not sure if Google is sending a nod or a wink... ;-)</p>
]]></description><pubDate>Tue, 10 Nov 2026 16:45:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=50442838</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50442838</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50442838</guid></item><item><title><![CDATA[New comment by beernet in "AlphaProof Code"]]></title><description><![CDATA[
<p>>> AlphaProof Code is a multi-LLM agentic framework that splits the work of formal verification the way a research team would<p>Not sure if this is precise. At least it sounds somewhat odd. I think what they really wanted to say here is "multi-agent system" (which usually, but not necessarily, implies the use of multiple LLMs).</p>
]]></description><pubDate>Tue, 10 Nov 2026 13:07:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=50439775</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50439775</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50439775</guid></item><item><title><![CDATA[New comment by beernet in "Dream-1"]]></title><description><![CDATA[
<p>So essentielly comments from simonw? Sounds like what you describe</p>
]]></description><pubDate>Sat, 24 Oct 2026 20:40:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=50225024</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50225024</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50225024</guid></item><item><title><![CDATA[New comment by beernet in "Timestral: Mistral release time-series foundation models"]]></title><description><![CDATA[
<p>You are telling me quant funds prefer a heavy-weight and slow ANN to a fast, specialized model? Maybe in totally non-critical environments, otherwise it's simply too slow, regardless of performance.</p>
]]></description><pubDate>Mon, 19 Oct 2026 12:03:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=50151354</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50151354</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50151354</guid></item><item><title><![CDATA[New comment by beernet in "Timestral: Mistral's time-series foundation models"]]></title><description><![CDATA[
<p>Meh....time series data is fundamentally different from textual data. I see why a time series "foundation model" seems appealing, especially to the marketing folks at Mistral. Real-world usability is very limited, which is probably the reason why the blog post features an imaginary customer rather than an actual reference.</p>
]]></description><pubDate>Mon, 19 Oct 2026 11:50:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=50151235</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50151235</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50151235</guid></item><item><title><![CDATA[New comment by beernet in "FLUX 3 Image Klein"]]></title><description><![CDATA[
<p>That's true. What's also true is that it cannot compete performance-wise.</p>
]]></description><pubDate>Sat, 17 Oct 2026 00:01:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=50126412</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50126412</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50126412</guid></item><item><title><![CDATA[New comment by beernet in "FLUX 3 Image Klein"]]></title><description><![CDATA[
<p>Yes, a prime example of when the performance vs "bias" tradeoff is handled anti-business</p>
]]></description><pubDate>Fri, 16 Oct 2026 23:59:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=50126392</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50126392</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50126392</guid></item><item><title><![CDATA[New comment by beernet in "3rd Yandex Cloud Data Center Was Hit"]]></title><description><![CDATA[
<p>> And of course not a word about the people in Russia killed by Ukranians (Kursk region and daily killings by drones, of civilians, which the west is "not aware of").<p>Russia could stop this invasion any time and not a single Russian would be hurt. Pure Kreml propaganda on HN, how sad.</p>
]]></description><pubDate>Sun, 11 Oct 2026 08:42:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=50040981</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=50040981</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50040981</guid></item><item><title><![CDATA[New comment by beernet in "German companies rely almost exclusively on AI models and services from the USA"]]></title><description><![CDATA[
<p>The short answer is No.<p>There are "sovereign" clouds like StackIT, OVH, OTC or Scaleway. What they all have in common is that they are lightyears behind hyperscalers in terms of performance, reliability and security. Needless to say, these alternatives can only serve open-source models because they are not capable of training a SOTA equivalent by themselves. The cherry on top is that you have to pay a sovereignty premium for inferior performance.</p>
]]></description><pubDate>Tue, 15 Sep 2026 14:59:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=49713574</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49713574</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49713574</guid></item><item><title><![CDATA[New comment by beernet in "Canada will match US tariffs 'dollar for dollar' as trade talks break down"]]></title><description><![CDATA[
<p>If you wanted a Russian puppet in the oval office, it would do exactly what Epstein's BFF Donald is doing. There are zero reasons to support anything Putin is doing (read: daily war crimes), however, the destruction from within in the US and EU is executed amazingly well.</p>
]]></description><pubDate>Sat, 22 Aug 2026 13:06:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=49399351</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49399351</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49399351</guid></item><item><title><![CDATA[New comment by beernet in "Turbovec – Google's TurboQuant for vector search in Rust"]]></title><description><![CDATA[
<p>Why not just use Qdrant? They've been integrating TurboQuant for months, works well.</p>
]]></description><pubDate>Tue, 18 Aug 2026 20:48:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=49352484</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49352484</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49352484</guid></item><item><title><![CDATA[New comment by beernet in "Mistral OCR 4.1"]]></title><description><![CDATA[
<p>Agreed. Does the GTM team there really sit together like "oh yeah, that sounds reasonable" while being totally beyond typical market prices?</p>
]]></description><pubDate>Thu, 13 Aug 2026 18:47:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=49290304</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49290304</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49290304</guid></item><item><title><![CDATA[New comment by beernet in "The Maxwell Conjecture Is False (GPT 5.6 Sol)"]]></title><description><![CDATA[
<p>On the one-hand side, it's really impressive how LLMs drive mathematics forward, and this pace is only accelerating very quickly.<p>At the same time, most of the proofs I've looked at appear super messy and chaotic to me (while still being correct of course, so it doesn't matter). LLMs do not care about "elegance" the way human beings do, which is a big advantage. LLMs for mathematics is such a great fit on many levels. Can't wait for a significant breakthrough, prove P=NP and all hell breaks loose.</p>
]]></description><pubDate>Fri, 31 Jul 2026 12:01:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=49122042</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49122042</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49122042</guid></item><item><title><![CDATA[New comment by beernet in "[dead]"]]></title><description><![CDATA[
<p>Just like the comment itself fwiw</p>
]]></description><pubDate>Wed, 29 Jul 2026 11:40:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=49096154</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49096154</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49096154</guid></item><item><title><![CDATA[New comment by beernet in "Hetzner is working on LLM Inference"]]></title><description><![CDATA[
<p>Paying a premium for inferior service quality is not exactly freedom</p>
]]></description><pubDate>Fri, 24 Jul 2026 21:40:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=49041931</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49041931</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49041931</guid></item><item><title><![CDATA[New comment by beernet in "Hetzner is working on LLM Inference"]]></title><description><![CDATA[
<p>To all mentions of IONOS / StackIT / OVH:<p>All of these are lightyears behind US/China LLM offerings. None of them offer any model close to open-source SOTA. Runtime and reliability is a disaster and sure enough you have to pay a "sovereignty" mark up.</p>
]]></description><pubDate>Fri, 24 Jul 2026 19:06:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=49040299</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=49040299</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49040299</guid></item><item><title><![CDATA[New comment by beernet in "Leanstral 1.5"]]></title><description><![CDATA[
<p>This went to market horribly (if you can even call it that), just look at the comments. Mistral played themselves big time over the past ~18 months. Non-competitive products and models combined with bad marketing and GTM...Oh Europe</p>
]]></description><pubDate>Wed, 01 Jul 2026 07:24:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=48743347</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=48743347</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48743347</guid></item><item><title><![CDATA[New comment by beernet in "Claude Sonnet 5"]]></title><description><![CDATA[
<p>> they have shit developer relations<p>Not true<p>> model that's mostly worse while being more expensive<p>Not true<p>> they can ride a wave of misinformation.<p>Not true</p>
]]></description><pubDate>Wed, 01 Jul 2026 07:20:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=48743319</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=48743319</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48743319</guid></item><item><title><![CDATA[New comment by beernet in "Claude Sonnet 5"]]></title><description><![CDATA[
<p>> Claude Code is still a meme<p>Claude Code generates more revenue than OpenAI...It appears to be a nice meme.</p>
]]></description><pubDate>Tue, 30 Jun 2026 19:33:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=48738096</link><dc:creator>beernet</dc:creator><comments>https://news.ycombinator.com/item?id=48738096</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48738096</guid></item></channel></rss>