<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: michaelchicory</title><link>https://news.ycombinator.com/user?id=michaelchicory</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sun, 04 Oct 2026 00:46:22 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=michaelchicory" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by michaelchicory in "Model Training as Code"]]></title><description><![CDATA[
<p>Savanna works in the same way as Nix: each training stage is defined fully in code, so it computes a hash over all inputs (training image, code version, environment, training configuration, etc.) and maps this to the eventual output. When the same inputs come along again, Savanna just reads the output from cache. As the post mentions, this is particularly helpful in sweeps that span multiple training stages.</p>
]]></description><pubDate>Mon, 29 Jun 2026 11:39:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=48717948</link><dc:creator>michaelchicory</dc:creator><comments>https://news.ycombinator.com/item?id=48717948</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48717948</guid></item><item><title><![CDATA[New comment by michaelchicory in "Model Training as Code"]]></title><description><![CDATA[
<p>Author here, happy to answer any questions!</p>
]]></description><pubDate>Mon, 29 Jun 2026 11:24:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=48717820</link><dc:creator>michaelchicory</dc:creator><comments>https://news.ycombinator.com/item?id=48717820</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48717820</guid></item><item><title><![CDATA[New comment by michaelchicory in "Show HN: Compile-Time Vibe Coding"]]></title><description><![CDATA[
<p>alas, even without this, that might be the case already :D</p>
]]></description><pubDate>Sat, 07 Feb 2026 16:48:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=46925263</link><dc:creator>michaelchicory</dc:creator><comments>https://news.ycombinator.com/item?id=46925263</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46925263</guid></item><item><title><![CDATA[New comment by michaelchicory in "Show HN: Compile-Time Vibe Coding"]]></title><description><![CDATA[
<p>I think it's cool that prompts are checked-in to version control, so the thought behind the action is captured. Also, there's a clear distinction between AI-generated and 'organic' code, which allows you to explicitly decouple core business logic from trivial grunt work. Although this particular implementation is likely not valuable, I feel there is value hiding somewhere in the concept. Certainly, the way humans currently collaborate with AI on code more often than not leads to unmaintainable software.</p>
]]></description><pubDate>Sat, 07 Feb 2026 16:43:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=46925215</link><dc:creator>michaelchicory</dc:creator><comments>https://news.ycombinator.com/item?id=46925215</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46925215</guid></item><item><title><![CDATA[Show HN: Compile-Time Vibe Coding]]></title><description><![CDATA[
<p>Worried about reproducible builds? Let OpenAI generate your source code at compile time.<p>Built mostly for the meme, but maybe there's something there...?</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46922238">https://news.ycombinator.com/item?id=46922238</a></p>
<p>Points: 11</p>
<p># Comments: 3</p>
]]></description><pubDate>Sat, 07 Feb 2026 08:19:56 +0000</pubDate><link>https://github.com/Michael-JB/vibecode</link><dc:creator>michaelchicory</dc:creator><comments>https://news.ycombinator.com/item?id=46922238</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46922238</guid></item><item><title><![CDATA[New comment by michaelchicory in "How ChatGPT spoiled my semester (2024)"]]></title><description><![CDATA[
<p>This sucks, though it’s formative. An experience that I value highly from my time studying was working on a group project in a team with misaligned goals: it teaches you how much it matters to find good people to work with in the real world!</p>
]]></description><pubDate>Thu, 07 Aug 2025 05:39:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=44820969</link><dc:creator>michaelchicory</dc:creator><comments>https://news.ycombinator.com/item?id=44820969</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44820969</guid></item></channel></rss>