<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: emaadm</title><link>https://news.ycombinator.com/user?id=emaadm</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 22:53:01 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=emaadm" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by emaadm in "Will I ever own a zettaflop?"]]></title><description><![CDATA[
<p>> We are already near the limits of what we can do<p>Hard disagree. If I had a million Claudes worth of compute I'd be livestreaming my entire reality feed to a local server 24/7 and having it organize my observations and thoughts, synthesize new ideas, implement prototypes and discard infeasible ones while I sleep. If you're in the business of knowledge creation, a million Claudes isn't enough. Text is an easy modality, I want foundation models that operate on text, images, audio, video, streaming point clouds, ...</p>
]]></description><pubDate>Fri, 10 Apr 2026 03:54:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=47713417</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=47713417</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47713417</guid></item><item><title><![CDATA[New comment by emaadm in "The turmoil at Thinking Machines Lab"]]></title><description><![CDATA[
<p>> Thinking Machines released a product that helps developers adjust A.I. models to better perform specific tasks. Offered to businesses and software developers over the internet, the product was similar to technologies from Google, Amazon and Microsoft.<p>I cannot take this article seriously.</p>
]]></description><pubDate>Fri, 23 Jan 2026 01:41:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=46727379</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=46727379</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46727379</guid></item><item><title><![CDATA[haggleforme.computer]]></title><description><![CDATA[
<p>Article URL: <a href="https://haggleforme.computer/">https://haggleforme.computer/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46524596">https://news.ycombinator.com/item?id=46524596</a></p>
<p>Points: 2</p>
<p># Comments: 1</p>
]]></description><pubDate>Wed, 07 Jan 2026 10:08:28 +0000</pubDate><link>https://haggleforme.computer/</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=46524596</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46524596</guid></item><item><title><![CDATA[New comment by emaadm in "Ask HN: What Are You Working On? (Nov 2025)"]]></title><description><![CDATA[
<p>Stopping Agents [1] - LLM agents that stop conversations early and save time.<p>Because they're trained using imitation learning instead of RL, they're scalable and easy to deploy with your own data (also open-source!).<p>Mainly targeted at and tested on quickly disqualifying prospects in sales calls, but can be applied more broadly.<p>[1] <a href="https://stoppingagents.com" rel="nofollow">https://stoppingagents.com</a></p>
]]></description><pubDate>Sun, 09 Nov 2025 22:52:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=45870047</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=45870047</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45870047</guid></item><item><title><![CDATA[Language agents for optimal conversation stopping]]></title><description><![CDATA[
<p>Article URL: <a href="https://stoppingagents.com/">https://stoppingagents.com/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=45712132">https://news.ycombinator.com/item?id=45712132</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Sun, 26 Oct 2025 14:22:57 +0000</pubDate><link>https://stoppingagents.com/</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=45712132</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45712132</guid></item><item><title><![CDATA[New comment by emaadm in "Announcing Tinker"]]></title><description><![CDATA[
<p>OpenAI has had a fine-tuning API since GPT-3.5, and a reinforcement fine-tuning API since last year.</p>
]]></description><pubDate>Thu, 02 Oct 2025 02:38:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=45445858</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=45445858</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45445858</guid></item><item><title><![CDATA[New comment by emaadm in "We reverse-engineered Flash Attention 4"]]></title><description><![CDATA[
<p>There's some interesting work in NeurIPS this year on fused kernels for MoE too: <a href="https://flash-moe.github.io/" rel="nofollow">https://flash-moe.github.io/</a></p>
]]></description><pubDate>Sun, 28 Sep 2025 15:40:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=45405145</link><dc:creator>emaadm</dc:creator><comments>https://news.ycombinator.com/item?id=45405145</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45405145</guid></item></channel></rss>