<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: criemen</title><link>https://news.ycombinator.com/user?id=criemen</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 09 Apr 2026 03:20:23 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=criemen" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by criemen in "Škoda DuoBell: A bicycle bell that penetrates noise-cancelling headphones"]]></title><description><![CDATA[
<p>Pretty cool if true!</p>
]]></description><pubDate>Wed, 08 Apr 2026 09:29:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=47687616</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47687616</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47687616</guid></item><item><title><![CDATA[New comment by criemen in "No, it doesn't cost Anthropic $5k per Claude Code user"]]></title><description><![CDATA[
<p>> What people don't realize is that cache is <i>free</i><p>I'm incredibly salty about this - they're essentially monetizing intensely something that allows them to sell their inference at premium prices to more users - without any caching, they'd have much less capacity available.</p>
]]></description><pubDate>Tue, 10 Mar 2026 09:08:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=47320758</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47320758</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47320758</guid></item><item><title><![CDATA[New comment by criemen in "Giving LLMs a personality is just good engineering"]]></title><description><![CDATA[
<p>I tried ChatGPT over the holidays (paid) vs. claude.ai (paid).
After trying some prompts that worked well on Claude in ChatGPT, I understand why people are so annoyed about AI slop. The speech patterns in text output for ChatGPT are both obvious and annoying, and impossible to unsee when people use them in written communication.<p>Claude isn't without problems ("You're absolutely right"), but I feel that some of the perception there is around the limited set of phrases the coding agent uses regularly, and comes less from the multi-paragraph responses from the chatbot.</p>
]]></description><pubDate>Wed, 04 Mar 2026 08:09:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=47244564</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47244564</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47244564</guid></item><item><title><![CDATA[New comment by criemen in "MacBook Air with M5"]]></title><description><![CDATA[
<p>> Out of curiosity, what are some good use cases for a MBP now with the MBAs being so powerful?<p>Local software development (node/TS). When opus-4.6-fast launched, it felt like some of the limiting factor in turnaround time moved from inference to the validation steps, i.e. execute tests, run linter, etc. Granted, that's with endpoint management slowing down I/O, and hopefully tsgo and some eslint replacement will speed things up significantly over there.</p>
]]></description><pubDate>Tue, 03 Mar 2026 19:06:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=47237153</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47237153</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47237153</guid></item><item><title><![CDATA[New comment by criemen in "Building SQLite with a small swarm"]]></title><description><![CDATA[
<p>Even if was copying sqlite code over, wouldn't the ability to automatically rewrite sqlite in Rust be a valuable asset?</p>
]]></description><pubDate>Mon, 16 Feb 2026 11:29:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=47033778</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47033778</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47033778</guid></item><item><title><![CDATA[New comment by criemen in "I want to wash my car. The car wash is 50 meters away. Should I walk or drive?"]]></title><description><![CDATA[
<p>> I had assumed that reasoning models should easily be able to answer this correctly.<p>I thought so too, yet Opus 4.6 with extended thinking (on claude.ai) gives me
> Walk. At 50 meters you'd spend more time parking and maneuvering at the car wash than the walk itself takes. Drive the car over only if the wash requires the car to be there (like a drive-through wash), then walk home and back to pick it up.<p>which is still pretty bad.</p>
]]></description><pubDate>Mon, 16 Feb 2026 11:21:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=47033724</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47033724</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47033724</guid></item><item><title><![CDATA[New comment by criemen in "I fixed Windows native development"]]></title><description><![CDATA[
<p>This is amazing.<p>At $workplace, we have a script that extracts a toolchain from a GitHub actions windows runner, packages it up, stuffs it into git LFS, which is then pulled by bazel as C++ toolchain.<p>This is the more scalable way, and I assume it could still somewhat easily be integrated into a bazel build.</p>
]]></description><pubDate>Sun, 15 Feb 2026 12:50:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=47023282</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47023282</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47023282</guid></item><item><title><![CDATA[New comment by criemen in "Two different tricks for fast LLM inference"]]></title><description><![CDATA[
<p>One other thing I'd assume Anthropic is doing is routing all fast requests to the latest-gen hardware. They most certainly have a diverse fleet of inference hardware (TPUs, GPUs of different generations), and fast will be only served by whatever is fastest, whereas the general inference workload will be more spread out.</p>
]]></description><pubDate>Sun, 15 Feb 2026 10:04:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=47022513</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47022513</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47022513</guid></item><item><title><![CDATA[Hamming, "You and Your Research" (1995) [video]]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.youtube.com/watch?v=a1zDuOPkMSw">https://www.youtube.com/watch?v=a1zDuOPkMSw</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47016112">https://news.ycombinator.com/item?id=47016112</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 14 Feb 2026 17:01:08 +0000</pubDate><link>https://www.youtube.com/watch?v=a1zDuOPkMSw</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=47016112</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47016112</guid></item><item><title><![CDATA[New comment by criemen in "Two Weeks Until Tapeout"]]></title><description><![CDATA[
<p>> aka: For those not living in 2026, we have uncovered a new clue to the mystery of where all the low-power DRAM chips have suddenly vanished to!<p>I love the writing style!</p>
]]></description><pubDate>Sun, 25 Jan 2026 12:03:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=46753353</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46753353</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46753353</guid></item><item><title><![CDATA[New comment by criemen in "Nvidia Kicks Off the Next Generation of AI with Rubin"]]></title><description><![CDATA[
<p>What's the power hookup to just boot one rack?
I'd imagine that's more than you get anywhere in residential areas for a single house.</p>
]]></description><pubDate>Thu, 08 Jan 2026 20:41:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=46546168</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46546168</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46546168</guid></item><item><title><![CDATA[New comment by criemen in "I switched from VSCode to Zed"]]></title><description><![CDATA[
<p>> I’m currently using a mix of Zed, Sublime, and VS Code.<p>Can you elaborate on when you use which editor?
I'd have imagined that there's value in learning and using one editor in-depth, instead of switching around based on use-case, so I'd love to learn more about your approach.</p>
]]></description><pubDate>Mon, 05 Jan 2026 23:05:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=46506414</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46506414</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46506414</guid></item><item><title><![CDATA[Is AI actually a Bubble?]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.newyorker.com/culture/open-questions/is-ai-actually-a-bubble">https://www.newyorker.com/culture/open-questions/is-ai-actually-a-bubble</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=46256759">https://news.ycombinator.com/item?id=46256759</a></p>
<p>Points: 3</p>
<p># Comments: 1</p>
]]></description><pubDate>Sat, 13 Dec 2025 18:32:57 +0000</pubDate><link>https://www.newyorker.com/culture/open-questions/is-ai-actually-a-bubble</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46256759</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46256759</guid></item><item><title><![CDATA[New comment by criemen in "Has the cost of building software dropped 90%?"]]></title><description><![CDATA[
<p>> This takes a fairly large mindset shift, but the hard work is the conceptual thinking, not the typing.<p>But the hard work always was the conceptual thinking? At least at and beyond the Senior level, for me it was always the thinking that's the hard work, not converting the thoughts into code.</p>
]]></description><pubDate>Mon, 08 Dec 2025 22:25:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=46198484</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46198484</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46198484</guid></item><item><title><![CDATA[New comment by criemen in "Has the cost of building software dropped 90%?"]]></title><description><![CDATA[
<p>The large open-weights models aren't really usable for local running (even with current hardware), but multiple providers compete on running inference for you, so it's reasonable to assume that there is and will be a functioning marketplace.</p>
]]></description><pubDate>Mon, 08 Dec 2025 22:14:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=46198364</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46198364</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46198364</guid></item><item><title><![CDATA[New comment by criemen in "The writing is on the wall for handwriting recognition"]]></title><description><![CDATA[
<p>It used to be that the answer was logprobs, but it seems that is no longer available.</p>
]]></description><pubDate>Wed, 03 Dec 2025 21:27:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=46140383</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46140383</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46140383</guid></item><item><title><![CDATA[New comment by criemen in "Hardening the C++ Standard Library at scale"]]></title><description><![CDATA[
<p>> Of course there is undefined behavior that isn't security critical.<p>But undefined behavior is literally introduced as "the compiler is allowed to do anything, including deleting all your files". Of course that's security critical by definition?</p>
]]></description><pubDate>Sat, 29 Nov 2025 21:16:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=46090849</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46090849</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46090849</guid></item><item><title><![CDATA[New comment by criemen in "It's Always the Process, Stupid"]]></title><description><![CDATA[
<p>I've been thinking a lot about that lately, and I agree. I used to be hard in the "You can't solve social problems with technical solutions", but that's not the whole truth.
If people aren't using your thing, sure, you can brand that as social problem (lack of buy-in on the process, people not being heard during rollout, ...). However one way of getting people to use your thing/process is to make it easier to use. Integrate it well into the workflow they're already familiar with, bring the tooling close, reduce friction, provide some extra value to your users with features etc.
That's technical solutions, but if you choose them based on knowledge of the "social problem" they can be quite effective.</p>
]]></description><pubDate>Sat, 29 Nov 2025 18:31:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=46089663</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46089663</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46089663</guid></item><item><title><![CDATA[New comment by criemen in ""Good engineering management" is a fad"]]></title><description><![CDATA[
<p>I've been thinking lately a lot about this. What is it I do when I want to convince someone of something (i.e. "creating alignment" in corporate speak)? I listen to them, am empathic, ask meaningful questions etc. Afterwards, that opens a space for me to make a proposal that is well-received.</p>
]]></description><pubDate>Sun, 23 Nov 2025 21:13:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=46027384</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=46027384</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46027384</guid></item><item><title><![CDATA[New comment by criemen in "Building more with GPT-5.1-Codex-Max"]]></title><description><![CDATA[
<p>Anthropic released the Opus 4.1 (basically, a new Opus 4 checkpoint) right around the big GPT-5 release date too, if I remember correctly. At this point, anything goes to stay relevant.</p>
]]></description><pubDate>Wed, 19 Nov 2025 22:50:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=45986398</link><dc:creator>criemen</dc:creator><comments>https://news.ycombinator.com/item?id=45986398</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45986398</guid></item></channel></rss>