<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: ryanschaefer</title><link>https://news.ycombinator.com/user?id=ryanschaefer</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sun, 21 Jun 2026 09:16:02 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=ryanschaefer" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by ryanschaefer in "Sam Bankman-Fried loses bid to appeal against fraud conviction in FTX case"]]></title><description><![CDATA[
<p>If Michael Lewis is to be believed making Gemini whole was his largest mistake.</p>
]]></description><pubDate>Sat, 13 Jun 2026 18:51:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=48520257</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48520257</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48520257</guid></item><item><title><![CDATA[New comment by ryanschaefer in "US ban on Mythos is related to a jailbreak research by Amazon researchers"]]></title><description><![CDATA[
<p>> According to a Wall Street Journal (WSJ) report, the technique in question was tested by researchers at Amazon,<p>Why not link that? Is this a dupe?<p>Found the primary source:<p><a href="https://www.wsj.com/tech/ai/amazon-ceos-talks-with-u-s-officials-triggered-crackdown-on-anthropic-models-dcc90578" rel="nofollow">https://www.wsj.com/tech/ai/amazon-ceos-talks-with-u-s-offic...</a></p>
]]></description><pubDate>Sat, 13 Jun 2026 17:48:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=48519617</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48519617</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48519617</guid></item><item><title><![CDATA[New comment by ryanschaefer in "H.R. 6028 would fundamentally change the U.S. Copyright Office"]]></title><description><![CDATA[
<p>> Members can request a division of the assembly (a rising vote, where each sides rise in turn to be counted)<p>Isn’t this the important bit? IIRC this can be demanded by anyone. If it passes by a voice vote, assume your representative voted yay or abstain.</p>
]]></description><pubDate>Sat, 13 Jun 2026 15:08:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=48518056</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48518056</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48518056</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Bot vs human traffic"]]></title><description><![CDATA[
<p>“First time”<p>The graph seems like it only goes back to April 27 and on that day it was 57% bot…</p>
]]></description><pubDate>Wed, 03 Jun 2026 18:12:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=48387579</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48387579</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48387579</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Don't Roll Your Own"]]></title><description><![CDATA[
<p>A strange issue I’ve found is careless use of AI at my job has lead to many people rolling their own incomplete mini parsers. Think YAML parsers of a frontmatter that expects either `key: value` exactly or treats `item1, item2, …` as a list.<p>It’s a litmus test I use to see if someone actually glanced over what the AI generated.</p>
]]></description><pubDate>Sun, 24 May 2026 16:19:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=48258526</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48258526</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48258526</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Amazon Web Services – Four Years and Out"]]></title><description><![CDATA[
<p>The problem is using AI to “push” the answer to an asker.
 Unless the company has hidden docs they use for support (which why would that ever benefit them), I could get just as good of an answer if I point my LLM at your docs (“pull” an answer). In fact, the response might be better because I have context set up to tune it to my understanding.<p>Instead, you (company / support agent) have decided that I should instead have a conversation with an LLM through a worse, more opaque harness to the detriment of all of us.</p>
]]></description><pubDate>Sun, 24 May 2026 14:14:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=48257445</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48257445</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48257445</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Flipper One – we need your help"]]></title><description><![CDATA[
<p>It’s the fat introduced by the process that annoys me the most. The user of the LLM had an idea, but it got greased up and packaged into something that the average person would create, not a specialist in the domain.
 It dumbs down everything into a single perspective / way of presenting a topic.</p>
]]></description><pubDate>Thu, 21 May 2026 13:01:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=48221950</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48221950</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48221950</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Flipper One – we need your help"]]></title><description><![CDATA[
<p>The oversized emdashes are <i>chefs kiss</i></p>
]]></description><pubDate>Thu, 21 May 2026 12:58:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=48221921</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48221921</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48221921</guid></item><item><title><![CDATA[New comment by ryanschaefer in "We're testing new ad formats in Search and expanding our Direct Offers pilot"]]></title><description><![CDATA[
<p>Entirely accurate, but what an absolute waste of resources across the board.</p>
]]></description><pubDate>Thu, 21 May 2026 12:09:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=48221345</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48221345</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48221345</guid></item><item><title><![CDATA[New comment by ryanschaefer in "How Claude Code works in large codebases"]]></title><description><![CDATA[
<p>In really large codebases grep and find timeout. If you operate at that scale you quickly come to realize Claude <i>will not</i> use the tools you built to make searching feasible.</p>
]]></description><pubDate>Fri, 15 May 2026 12:27:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=48147775</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48147775</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48147775</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Rewrite Bun in Rust has been merged"]]></title><description><![CDATA[
<p>It’s a watershed moment. Basically one of the most controlled applications of an LLM into a robust codebase without regard for the implications of doing so.<p>Anthropic needed something like this and it must proceed flawlessly. My guess is that nothing will <i>explicitly</i> break. But that’s the difficulty of LLM generated code: nothing breaks. You sit with a codebase that swallows all errors and appears to be working. Silently failing makes debugging performance and behavior much harder.</p>
]]></description><pubDate>Thu, 14 May 2026 19:47:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=48140262</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48140262</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48140262</guid></item><item><title><![CDATA[New comment by ryanschaefer in "[dead]"]]></title><description><![CDATA[
<p>> Is this your post? We don't fall asleep as easily for Write.as subscribers.</p>
]]></description><pubDate>Thu, 14 May 2026 16:10:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=48137433</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48137433</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48137433</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Rewrite Bun in Rust has been merged"]]></title><description><![CDATA[
<p>I think one of the things I had forgotten about but sheds some more light in my mind about <i>how</i> this was done is that anthropic bought bun.<p>The change of tone with the author in the capabilities of Claude. The strategy of merging everything at once instead of a more slow, careful cutover. The “single” author story that every company loves to put forth.</p>
]]></description><pubDate>Thu, 14 May 2026 15:32:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=48136934</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48136934</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48136934</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Postmortem: TanStack NPM supply-chain compromise"]]></title><description><![CDATA[
<p>> many devs would prefer convenience and tradition<p>This is too reductive of the situation.<p>If it ain’t broke don’t fix it. Except, in this case, unless you have someone tell you it’s broken you won’t even know you need to fix it.<p>And this is where asymmetry comes in to play. Attackers are free to test and break as much as they want as long as they are silent. Whereas maintainers don’t know if the fix an LLM proposes will actually address the issue or cause some regression elsewhere.<p>IMO, if Microsoft wants actually good PR around GitHub for once they would offer free LLM security audits on all actions for at least the X most popular repos…</p>
]]></description><pubDate>Tue, 12 May 2026 11:00:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=48106538</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48106538</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48106538</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Postmortem: TanStack NPM supply-chain compromise"]]></title><description><![CDATA[
<p><a href="https://noyaml.com" rel="nofollow">https://noyaml.com</a></p>
]]></description><pubDate>Tue, 12 May 2026 10:22:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=48106281</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48106281</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48106281</guid></item><item><title><![CDATA[New comment by ryanschaefer in "The React2Shell Story"]]></title><description><![CDATA[
<p>> But that afternoon, fueled by curiosity and frustration, I felt a switch flip in my brain, and I dived head-first into a rabbit hole with no turning back.<p><a href="https://xkcd.com/356/" rel="nofollow">https://xkcd.com/356/</a><p>It happens to all of us. However, I think it’s much easier nowadays with LLMs for something productive like this to come out of it. I can notice something wrong and triage or even fix it before the point where I’d normally start to feel the subconscious pull of opportunity cost telling me to stop.</p>
]]></description><pubDate>Sat, 09 May 2026 15:16:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=48075651</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48075651</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48075651</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Tesla is recalling its cheaper Cybertruck because the wheels might fall off"]]></title><description><![CDATA[
<p>We’ve taken it *outside* the environment</p>
]]></description><pubDate>Fri, 08 May 2026 15:05:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=48064248</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48064248</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48064248</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Become a Claude Certified Architect"]]></title><description><![CDATA[
<p>Are we even at the stage where we can encode <i>any</i> best practices into such a certification…?</p>
]]></description><pubDate>Fri, 08 May 2026 15:02:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=48064208</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=48064208</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48064208</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Claude Opus 4.7 costs 20–30% more per session"]]></title><description><![CDATA[
<p>What?</p>
]]></description><pubDate>Fri, 17 Apr 2026 21:16:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=47810658</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=47810658</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47810658</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Show HN: Is Hormuz open yet?"]]></title><description><![CDATA[
<p>For me, I had a similar reaction to the original commenter.<p>I thought about it more and realized it’s the “via Polymarket” tag on it. It reads like an advertisement. I don’t need to know the exact market the figure is from. Someone could even combine multiple ones and I still wouldn’t want to know.</p>
]]></description><pubDate>Thu, 09 Apr 2026 12:01:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=47702544</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=47702544</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47702544</guid></item></channel></rss>