<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>Wed, 08 Apr 2026 10:28:19 +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 "Caveman: Why use many token when few token do trick"]]></title><description><![CDATA[
<p>Kinda ironic this description is so verbose.<p>> Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman<p>For the first part of this: couldn’t this just be a UserSubmitPrompt hook with regex against these?<p>See additionalContext in the json output of a script: <a href="https://code.claude.com/docs/en/hooks#structured-json-output" rel="nofollow">https://code.claude.com/docs/en/hooks#structured-json-output</a><p>For the second, /caveman will always invoke the skill /caveman: <a href="https://code.claude.com/docs/en/skills" rel="nofollow">https://code.claude.com/docs/en/skills</a></p>
]]></description><pubDate>Sun, 05 Apr 2026 10:23:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=47647955</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=47647955</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47647955</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Universal Claude.md – cut Claude output tokens"]]></title><description><![CDATA[
<p>The whole “Code Output” section is horrifying especially with how I have seen Claude operate in a large monorepo.<p>This mode of operation results in hacks on top of shaky hacks on top of even flimsier, throw away, absolutely sloppy hacks.<p>An example - using dict like structs instead of classes. Claude really likes to load all of the data that it can aggressively even if it’s not needed. This further exhibits itself as never wanting to add something directly <i>to</i> a class and instead wanting to add <i>around</i> it.</p>
]]></description><pubDate>Tue, 31 Mar 2026 05:08:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=47582960</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=47582960</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47582960</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Apple Business"]]></title><description><![CDATA[
<p>> And Apple Business can help millions of companies grow their reach and connect with local customers across Apple Maps, Mail, Wallet, Siri, and more, including a new option coming this summer that will enable businesses in the U.S. and Canada to place local ads in Maps during key search and discovery moments. Apple Business will be available starting Tuesday, April 14, in more than 200 countries and regions<p>Burying the lede about ADs coming to <i>everything</i> in this announcement. Seems like the contract most people implicitly signed when choosing Apple just broke.</p>
]]></description><pubDate>Wed, 25 Mar 2026 11:48:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=47516066</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=47516066</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47516066</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Show HN: Mcp2cli – One CLI for every API, 96-99% fewer tokens than native MCP"]]></title><description><![CDATA[
<p>> Per turn<p>Isn’t this somewhat misleading? Any system context is going to be added “per turn” because it’s included in the first turn.<p>Is any context removed on a turn by turn basis (aside from thinking?)</p>
]]></description><pubDate>Tue, 10 Mar 2026 13:17:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=47322877</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=47322877</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47322877</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Tiny sugar spoons are popping up on NYC fast-food menus"]]></title><description><![CDATA[
<p>He’s back making that exact product now</p>
]]></description><pubDate>Wed, 22 Oct 2025 15:01:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=45670205</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=45670205</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45670205</guid></item><item><title><![CDATA[New comment by ryanschaefer in "The Rapper 50 Cent, Adjusted for Inflation"]]></title><description><![CDATA[
<p>Made this a while back riffing on the same idea<p><a href="https://ryanschaefer.github.io/Its-one-banana-michael/" rel="nofollow">https://ryanschaefer.github.io/Its-one-banana-michael/</a></p>
]]></description><pubDate>Fri, 17 Oct 2025 21:52:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=45622550</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=45622550</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45622550</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Prompts for our Grok chat assistant and grok bot on X"]]></title><description><![CDATA[
<p>Call me cynical but this smells like a PR stunt to get people talking about how much controlling the system prompt matters to everyone’s interaction with assistants.<p>Most people wouldn’t give it a second thought unless something like this happened and it’s now plain to see how quickly they can be (poorly) manipulated.</p>
]]></description><pubDate>Fri, 16 May 2025 13:36:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=44005375</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=44005375</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44005375</guid></item><item><title><![CDATA[New comment by ryanschaefer in "A $20k American-made electric pickup with no paint, no stereo, no screen"]]></title><description><![CDATA[
<p>So incredibly refreshing after the abomination that is the telos was announced</p>
]]></description><pubDate>Sat, 26 Apr 2025 02:56:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=43800498</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=43800498</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43800498</guid></item><item><title><![CDATA[New comment by ryanschaefer in "ChatGPT 4.1 Jailbreak Prompt"]]></title><description><![CDATA[
<p>I’m sorry… what?</p>
]]></description><pubDate>Wed, 16 Apr 2025 16:10:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=43707125</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=43707125</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43707125</guid></item><item><title><![CDATA[New comment by ryanschaefer in "API design note: Beware of adding an "Other" enum value"]]></title><description><![CDATA[
<p>> Fewer moving pieces == less chance for bad times.<p>Is this not a case for explicitly specifying all flavors? Other flavor has essentially introduced infinite moving pieces.</p>
]]></description><pubDate>Sun, 02 Mar 2025 22:33:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=43235971</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=43235971</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43235971</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Argentina president faces impeachment calls over crypto crash"]]></title><description><![CDATA[
<p>Some were, but then again Ethereum was an ICO as was XRP.<p>Again with DeFi, some were just Ponzi schemes (see SBFs infamous box token comments) but it also gave us Uniswap and Compound. These at least allow you to take directional bets on crypto so you can make money on its downfall.<p>> and they can't keep up with the rate of new scam technology that comes out of crypto.<p>There’s really been no innovation in scams recently. Just highly promote a smart contract token you can buy early on then dump it.<p>> but a lot of them are effectively ICOs under a different name.<p>They explicitly cannot promise to do anything economically productive with the offering because that fails the Howey test.</p>
]]></description><pubDate>Wed, 19 Feb 2025 13:41:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=43102039</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=43102039</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43102039</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Argentina president faces impeachment calls over crypto crash"]]></title><description><![CDATA[
<p>It’s interesting. I remember hearing something I agree with from one of Bloomberg’s podcasts (either Moneystuff or Odd Lots) that the ambiguous regulatory regime for crypto actually lead to the state we’re in.<p>Summarizing, but what they generally said is that the actual “disruptive” attempts at technology from crypto (ICOs / DeFi) were selectively but heavily punished. Take XRP for an example of that.<p>What wasn’t / isn’t punished is meme coins because they fail the Howey test and there’s not really anything that a body which regulates investments can do against something screaming “we are not an investment.”</p>
]]></description><pubDate>Tue, 18 Feb 2025 14:58:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=43090237</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=43090237</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43090237</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Is the TikTok ban a chance to rethink the whole internet?"]]></title><description><![CDATA[
<p>But larger creators get better rates from the likes of Twitch and YouTube.</p>
]]></description><pubDate>Mon, 20 Jan 2025 02:40:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=42764433</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=42764433</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42764433</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Is the TikTok ban a chance to rethink the whole internet?"]]></title><description><![CDATA[
<p>> Advertisers love that, and so do platform owners. Its much easier to control and squeeze a few creators rather than a big diffuse group.<p>There’s most likely some middle ground for this. Too diffuse and it makes direct sponsorship difficult meaning the platform has more control over advertisements as they need to be algorithmically placed. Too centralized and you’re back in the land of cable where your advertising is less effective because of the broad cross section of the audience watching.<p>As far as “squeezing” creators goes… no, more centralized means higher demand which means higher pay and more control in creators hands.</p>
]]></description><pubDate>Sat, 18 Jan 2025 13:34:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=42748206</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=42748206</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42748206</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Data evolution with set-theoretic types"]]></title><description><![CDATA[
<p>> what you want to do is easy
 Easy to implement, hard to get correct. It inverts where you do work in a system.
It can be hard to implement robust types but once that’s done it’s easy to know what you are writing is correct.</p>
]]></description><pubDate>Tue, 14 Jan 2025 19:27:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=42702473</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=42702473</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42702473</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Niantic announces “Large Geospatial Model” trained on Pokémon Go player data"]]></title><description><![CDATA[
<p>Who knows what is next</p>
]]></description><pubDate>Thu, 21 Nov 2024 04:30:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=42201134</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=42201134</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42201134</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Niantic announces “Large Geospatial Model” trained on Pokémon Go player data"]]></title><description><![CDATA[
<p>I’m not exactly shocked that it could exist. But this usage (beyond the scope of processing barcodes) seems like it couldn’t be construed to fit into the normal avenues of data collection under a privacy policy.
 Also with regard to training specifically, this policy was created in late 2020 so I don’t know how it would cover generative models.</p>
]]></description><pubDate>Thu, 21 Nov 2024 04:09:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=42201014</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=42201014</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42201014</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Niantic announces “Large Geospatial Model” trained on Pokémon Go player data"]]></title><description><![CDATA[
<p>I’d be interested in how your privacy policy allows this. I can’t find where it mentions photos are stored or used for training purposes…</p>
]]></description><pubDate>Thu, 21 Nov 2024 03:59:23 +0000</pubDate><link>https://news.ycombinator.com/item?id=42200951</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=42200951</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42200951</guid></item><item><title><![CDATA[New comment by ryanschaefer in "Remember That DNA You Gave 23andMe?"]]></title><description><![CDATA[
<p>What laws are present in the EU that individuals in US could look to to write to our representatives about?</p>
]]></description><pubDate>Sat, 28 Sep 2024 14:25:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=41680408</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=41680408</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41680408</guid></item><item><title><![CDATA[New comment by ryanschaefer in "An ALS drug fails, again"]]></title><description><![CDATA[
<p>Could you explain more what you mean by this?<p>I guess there’s the fact it made it to Phase III before they canned it meaning more people were hopeful than if action were taken to stop progress sooner.<p>But, how can progress be made if failure is disallowed because of unknowingly giving people false hope?</p>
]]></description><pubDate>Sat, 09 Mar 2024 17:16:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=39653107</link><dc:creator>ryanschaefer</dc:creator><comments>https://news.ycombinator.com/item?id=39653107</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39653107</guid></item></channel></rss>