<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: abtinf</title><link>https://news.ycombinator.com/user?id=abtinf</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 30 May 2026 21:16:55 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=abtinf" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by abtinf in "SQLite is all you need for durable workflows"]]></title><description><![CDATA[
<p>There are many cases where SQLite + concurrent front end (like a go net/http server) can handle all the load that a service might ever conceivably have to handle, especially if allowed to scale up hardware over time. You can trivially scale up SQLite to, what, hundreds of thousands of tps?<p>The only thing you really give up is HA/failover and DR. But there are solutions to deal with those. And single-server systems are generally surprisingly robust (since, in the absence of very complex control planes, uptime goes down with more systems).</p>
]]></description><pubDate>Fri, 29 May 2026 19:37:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=48328167</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48328167</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48328167</guid></item><item><title><![CDATA[New comment by abtinf in "Ferrari Luce"]]></title><description><![CDATA[
<p>> Pontiac Aztek was at least bold<p>How bad does a design have to be that this is a valid attack?<p>The Luce is so generic it borders on nihilism - destroying the very concept of Ferrari precisely because Ferraris are good.</p>
]]></description><pubDate>Tue, 26 May 2026 16:05:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=48281623</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48281623</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48281623</guid></item><item><title><![CDATA[New comment by abtinf in "Show HN: Write your BPF programs in Go, not C"]]></title><description><![CDATA[
<p>This transpiles to the native language.</p>
]]></description><pubDate>Mon, 25 May 2026 22:21:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=48272605</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48272605</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48272605</guid></item><item><title><![CDATA[New comment by abtinf in "Toxic chemical leak at a manufacturing facility in Orange County"]]></title><description><![CDATA[
<p>That area has dozens of aerospace manufacturers, building up since before WW2. People wanted to live close to work. There are lots of homes and commercial areas and industrial parks are tightly mixed together.<p>Source: I’ve worked in aerospace in Orange County.</p>
]]></description><pubDate>Sat, 23 May 2026 23:12:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=48252527</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48252527</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48252527</guid></item><item><title><![CDATA[New comment by abtinf in "The current AI pricing was always going to go away"]]></title><description><![CDATA[
<p>I don’t see how models can be licensed at all. There is no creative element in them.<p>As you say, you start with a random array and start mutating it until you get something that magically does interesting things.<p>Sure, you can hold copyright over all the software used to train the thing. And trade secrets or patents around your data selection, training methods, and infrastructure and such.<p>But unlike typical software compilation, the model isn’t a rote translation of something that has a creative element. Ordinary software has creative source code as input, mechanically processed into an output.<p>Models start with a bunch of inputs that are not the creative property of the model maker. Those non-creative inputs are not imbued with novel creativity, no matter how advanced the intermediate machinery may be.<p>By analogy, you may hold a copyright on the layout and creative elements of a phone book, but you have no rights over the actual data of phone numbers. Nor will any amount of ingenious layout engines or ad placement algorithms or complex printing press methods turn those numbers into something that can be licensed.<p>IANAL. This is truly baffling to me and it seems like everyone is going along with it because some corporate lawyer probably said “Iunno, let’s just say we are licensing this thing before release. Worst case, a court throws out the license”.</p>
]]></description><pubDate>Fri, 22 May 2026 22:36:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=48242496</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48242496</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48242496</guid></item><item><title><![CDATA[New comment by abtinf in "The current AI pricing was always going to go away"]]></title><description><![CDATA[
<p>Insofar as I can tell, inference is on a certain path toward becoming "free". The models are now extremely powerful on high-end consumer hardware, and the efficiency trend seems likely to continue.<p>Here is a recent non-rigorous benchmark I ran against a bunch of models. Qwen3.6 35B A3B fine-tuned with opus data runs plenty fast on my local machine and produce outstanding results - easily in the top 5, comparable to GPT 5.5 Pro (which is $180/mtok).<p><a href="https://gistpreview.github.io/?31d66ef69e4aed3efae1aec69d86c298/" rel="nofollow">https://gistpreview.github.io/?31d66ef69e4aed3efae1aec69d86c...</a><p>I've predicted for years now that the industry will head down the path of the virus scanning vendors: selling subscriptions to be able to download the latest versions of models. I simply don't see how any other business model is remotely viable, except at the very highest end of inference or video gen.</p>
]]></description><pubDate>Fri, 22 May 2026 15:47:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=48237462</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48237462</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48237462</guid></item><item><title><![CDATA[New comment by abtinf in "Using Kagi Search with Low Vision"]]></title><description><![CDATA[
<p>This approach ignores 99% of the value of Kagi. Google provides atrocious results compared to Kagi once you've taken a few minutes to use the basic Kagi features.</p>
]]></description><pubDate>Fri, 22 May 2026 13:37:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=48235689</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48235689</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48235689</guid></item><item><title><![CDATA[New comment by abtinf in "Not alive, but not dead: disembodied human brains used for drug testing"]]></title><description><![CDATA[
<p>Your doubts are solidly founded; I can confirm I am not yet dead and my organs have not yet been donated.</p>
]]></description><pubDate>Thu, 21 May 2026 02:17:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=48217008</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48217008</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48217008</guid></item><item><title><![CDATA[New comment by abtinf in "Not alive, but not dead: disembodied human brains used for drug testing"]]></title><description><![CDATA[
<p>I will be removing my organ donor status. This is horrifying.</p>
]]></description><pubDate>Wed, 20 May 2026 20:55:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=48214018</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48214018</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48214018</guid></item><item><title><![CDATA[New comment by abtinf in "Gemini 3.5 Flash"]]></title><description><![CDATA[
<p>hesamation/Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-GGUF @ Q6_K<p>8112 tokens @ 52.97 TPS, 0.85s TTFT<p><a href="https://gistpreview.github.io/?7bdefff99aca89d1bc12405323bd4da4" rel="nofollow">https://gistpreview.github.io/?7bdefff99aca89d1bc12405323bd4...</a><p>Full session: <a href="https://gist.github.com/abtinf/7bdefff99aca89d1bc12405323bd4da4#file-frog-rowing-jungle-river-2026-05-19-11-33-md" rel="nofollow">https://gist.github.com/abtinf/7bdefff99aca89d1bc12405323bd4...</a><p>Generated with LM Studio on a Macbook Pro M2 Max<p><a href="https://huggingface.co/hesamation/Qwen3.6-35B-A3B-Claude-4.6-Opus-Reasoning-Distilled-GGUF" rel="nofollow">https://huggingface.co/hesamation/Qwen3.6-35B-A3B-Claude-4.6...</a></p>
]]></description><pubDate>Tue, 19 May 2026 18:42:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=48197486</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48197486</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48197486</guid></item><item><title><![CDATA[New comment by abtinf in "Homunctor – The Simplest Agent"]]></title><description><![CDATA[
<p>I made a repo showing the smallest agent I could come up with.<p>If you're like me, the "magic" around agents is uncomfortable. I needed to understand precisely how they work. Partly inspired by the simplicity of the Pi agent harness, I decided to build the simplest agent I possibly could.<p>I also made a video walkthrough building it up, from a simple console agent to a self-modifying harness, in a single page HTML file.</p>
]]></description><pubDate>Mon, 18 May 2026 14:26:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=48180383</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48180383</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48180383</guid></item><item><title><![CDATA[Homunctor – The Simplest Agent]]></title><description><![CDATA[
<p>Article URL: <a href="https://github.com/abtinf/homunctor">https://github.com/abtinf/homunctor</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48180382">https://news.ycombinator.com/item?id=48180382</a></p>
<p>Points: 1</p>
<p># Comments: 1</p>
]]></description><pubDate>Mon, 18 May 2026 14:26:33 +0000</pubDate><link>https://github.com/abtinf/homunctor</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48180382</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48180382</guid></item><item><title><![CDATA[New comment by abtinf in "Eric Schmidt speech about AI booed during graduation"]]></title><description><![CDATA[
<p>Shouting people down isn’t “open debate”.</p>
]]></description><pubDate>Mon, 18 May 2026 13:25:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=48179543</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48179543</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48179543</guid></item><item><title><![CDATA[New comment by abtinf in "Driver accused of DUI tracks missing laptop to Illinois State trooper's house"]]></title><description><![CDATA[
<p>A field sobriety test is distinct from a chemical analysis (breathalyzer or otherwise).<p>In California, you are required to submit to chemical testing (breath, urine, or blood — I don’t recall the rules for which applies in which situations). However, you are not required to otherwise talk to or perform the absurd procedure of the field sobriety test (“you have the right to remain silent”).</p>
]]></description><pubDate>Mon, 11 May 2026 13:58:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=48095108</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48095108</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48095108</guid></item><item><title><![CDATA[New comment by abtinf in "The FCC wants your ID before you get a phone number"]]></title><description><![CDATA[
<p>> there should be verifiable reverse lookup to a valid registered number<p>This has existed for a very, very long time; its just a business feature you have to pay a lot of extra money for and that is generally unavailable to consumers.<p>Business customers get access to DNIS and ANI.<p>DNIS is the number that was dialed to reach the business. It is used for call routing and marketing analytics.<p>ANI is the phone number from which the call was made. It’s used again for call routing and analytics, but also billing. It’s how they knew where to send the bill back in the days of 1-900 numbers. Because real money is involved, it’s a pretty good bet it’s usually tied to the real entity.<p>You, as a mere consumer, are left with caller id, which no business user cares about.</p>
]]></description><pubDate>Sat, 09 May 2026 18:54:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=48077276</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48077276</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48077276</guid></item><item><title><![CDATA[New comment by abtinf in "Bun is being ported from Zig to Rust"]]></title><description><![CDATA[
<p>That is a devastating comment. I will now be extremely skeptical of bun.</p>
]]></description><pubDate>Tue, 05 May 2026 03:49:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=48017844</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=48017844</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48017844</guid></item><item><title><![CDATA[New comment by abtinf in "An update on recent Claude Code quality reports"]]></title><description><![CDATA[
<p>You didn’t anticipate most people stick with defaults?</p>
]]></description><pubDate>Thu, 23 Apr 2026 22:41:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=47883159</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=47883159</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47883159</guid></item><item><title><![CDATA[New comment by abtinf in "SpaceX says it has agreement to acquire Cursor for $60B"]]></title><description><![CDATA[
<p>You can just reallocate away from an index fund.</p>
]]></description><pubDate>Wed, 22 Apr 2026 01:11:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=47857394</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=47857394</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47857394</guid></item><item><title><![CDATA[New comment by abtinf in "SpaceX says it has agreement to acquire Cursor for $60B"]]></title><description><![CDATA[
<p>Nasdaq is an exchange. S&P 500 is an index.<p>S&P 500 includes companies from multiple exchanges. Like Nvidia, which lists on Nasdaq.</p>
]]></description><pubDate>Wed, 22 Apr 2026 01:10:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=47857381</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=47857381</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47857381</guid></item><item><title><![CDATA[New comment by abtinf in "Claude Code to be removed from Anthropic's Pro plan?"]]></title><description><![CDATA[
<p>Perhaps vibe coding the A/B testing engine isn't the best idea.</p>
]]></description><pubDate>Wed, 22 Apr 2026 00:57:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=47857255</link><dc:creator>abtinf</dc:creator><comments>https://news.ycombinator.com/item?id=47857255</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47857255</guid></item></channel></rss>