<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: km144</title><link>https://news.ycombinator.com/user?id=km144</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 21:42:41 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=km144" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by km144 in "Garage Door Spam"]]></title><description><![CDATA[
<p>The recommended videos are also definitely ads. There is no way there's no payola going on here.</p>
]]></description><pubDate>Thu, 19 Nov 2026 21:24:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=50566590</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=50566590</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50566590</guid></item><item><title><![CDATA[New comment by km144 in "Jensen Huang resigns"]]></title><description><![CDATA[
<p>I'm doubtful of this hypothesis. Maybe for a few years, sure. But China has proven time and again that they do not need foreign supply constraints to incentivize the growth of their own competing domestic industry.<p>The fundamental contradiction of the AI maximalists like Huang is claiming that AI does not need regulated while simultaneously supporting chip exports to China. The supporting claim for the former is that we need to compete on AI with China and move as fast as possible and every day matters, while the latter implies that slowing down their compute is inconsequential.</p>
]]></description><pubDate>Thu, 12 Nov 2026 18:20:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=50471716</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=50471716</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50471716</guid></item><item><title><![CDATA[New comment by km144 in "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"]]></title><description><![CDATA[
<p>The time between public announcement -> public release seems to be a tacit admission that Google is still a bit behind the frontier labs. Google is justifying the delay by pointing to cyber risk and a closed-access period reserved for analysis of safeguards. I'm dubious of this justification. But if it is true, I'm pretty sure the other labs are doing similar sorts of things in private. I think most of us prefer it that way—if Anthropic public announced Fable 5.5 two weeks ago with no availability that would add to my annoyance. Maybe they learned their lesson with Mythos, or  maybe it's just that Google also does not have an IPO looming. Who knows?</p>
]]></description><pubDate>Thu, 29 Oct 2026 14:31:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=50289725</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=50289725</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50289725</guid></item><item><title><![CDATA[New comment by km144 in "Every AI model has the same "controversial" take on humans"]]></title><description><![CDATA[
<p>Most people I run into seem far less sensitive to the bad writing of Claude than I am (and I assume you share this). For most people, they were satisfied enough with the improvements in Opus 5.5 to finally accept the prose. I still find it wildly unacceptable for models that are so good at writing software. After all, language is a pretty important part of writing software. This writing is very bad. I tolerate reading it for work but I cannot tolerate reading an article written by these models.</p>
]]></description><pubDate>Fri, 23 Oct 2026 15:14:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=50210464</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=50210464</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50210464</guid></item><item><title><![CDATA[New comment by km144 in "Claude Opus 5.5"]]></title><description><![CDATA[
<p>I think this release is really going to give them a hard time selling Fable:<p>> On our benchmarks, Claude Opus 5.5 leads in agentic coding, computer use, and knowledge work. That said, at these levels of capability we’ve found that benchmark margins have become a less reliable guide to real-world differences. In our own use, the gap between Opus 5.5 and Claude Fable 5.1 is narrower than these scores suggest.<p>In general, "benchmark margins have become a less reliable guide to real-world differences" sounds like a big problem. It was certainly the biggest problem with the previous generation of Claude models for a different reason, because the non-code output was nonsensical, and that is not being benchmarked at the moment. But I'm not sure what to make of this admission.</p>
]]></description><pubDate>Tue, 22 Sep 2026 16:41:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=49804122</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49804122</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49804122</guid></item><item><title><![CDATA[New comment by km144 in "Claude Opus 5.5"]]></title><description><![CDATA[
<p>Can you fix the link on <i>that</i> post then? I duped because that post links to a diff that tells me nothing about Opus 5.5</p>
]]></description><pubDate>Tue, 22 Sep 2026 16:35:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=49803995</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49803995</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49803995</guid></item><item><title><![CDATA[Claude Opus 5.5]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.anthropic.com/claude-opus-5-5">https://www.anthropic.com/claude-opus-5-5</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49803892">https://news.ycombinator.com/item?id=49803892</a></p>
<p>Points: 1806</p>
<p># Comments: 1134</p>
]]></description><pubDate>Tue, 22 Sep 2026 16:29:05 +0000</pubDate><link>https://www.anthropic.com/claude-opus-5-5</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49803892</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49803892</guid></item><item><title><![CDATA[New comment by km144 in "A coffee shop owner used AI to make a menu poster. Then came the angry DMs"]]></title><description><![CDATA[
<p>When I read that, the only thing I could think was "holy shit get off your fucking phone". But I agree with you as well.</p>
]]></description><pubDate>Thu, 17 Sep 2026 12:25:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=49739767</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49739767</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49739767</guid></item><item><title><![CDATA[New comment by km144 in "Astra for Coding: Why Are We Doing This Again?"]]></title><description><![CDATA[
<p>The biggest jump was Opus 4.6. Since then they have gradually gotten better at finding issues in your reasoning, not hallucinating, and being rigorous with the code, but much much worse at explaining things and generally just talking in a way that a human can understand. All the models I've tried seem to be suffering from the same fate so it must be something going on with the training meta right now.</p>
]]></description><pubDate>Fri, 11 Sep 2026 19:45:02 +0000</pubDate><link>https://news.ycombinator.com/item?id=49664299</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49664299</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49664299</guid></item><item><title><![CDATA[New comment by km144 in "Claude Fable 5.1 and Claude Mythos 5.1"]]></title><description><![CDATA[
<p>Yep. Simple answer is they want to IPO in the fall, and a new Haiku does literally nothing for them</p>
]]></description><pubDate>Tue, 01 Sep 2026 18:37:23 +0000</pubDate><link>https://news.ycombinator.com/item?id=49526102</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49526102</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49526102</guid></item><item><title><![CDATA[New comment by km144 in "Fastpotify"]]></title><description><![CDATA[
<p>That would actually be a reasonable thing to do with a model that is good at writing though. I'm sure if you used that method on Opus 4.6 you'd get a pretty decent result, because Claude models used to be quite good at that sort of technical synthesis. Asking a model that is bad at technical writing (e.g. Opus/Fable 5) to write your docs is a bad idea.</p>
]]></description><pubDate>Tue, 01 Sep 2026 14:14:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=49522314</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49522314</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49522314</guid></item><item><title><![CDATA[New comment by km144 in "Fastpotify"]]></title><description><![CDATA[
<p>I believe AI technical prose peaked with Opus 4.6. I still use it (mostly for that purpose) and I think it's legitimately great. I'm hoping they are able to reverse the trends that benchmaxxing and RLAIF have wrought on Claude's non-code output with the future models.</p>
]]></description><pubDate>Tue, 01 Sep 2026 14:11:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=49522261</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49522261</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49522261</guid></item><item><title><![CDATA[New comment by km144 in "OpenLogi"]]></title><description><![CDATA[
<p>I assure you it won't be as bad as this (answer to FAQ: "Will OpenLogi support Logitech Flow?")<p>> It's on the roadmap, at the far end: a cross-computer pointer and clipboard bridge is a very large feature. The half that lives in the protocol already ships. OpenLogi drives Easy-Switch host switching over HID++ (0x1814/0x1815), and paired mice follow the keyboard when it switches hosts. If the rest lands, it will be opt-in and local-network only.<p>A great example of why Claude's writing is terrible is the sentence "The half that lives in the protocol already ships". I cannot imagine any human writing this.</p>
]]></description><pubDate>Wed, 19 Aug 2026 16:36:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=49363809</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49363809</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49363809</guid></item><item><title><![CDATA[New comment by km144 in "Qwen3.8 Max now ranked as the best overall model by agentic index"]]></title><description><![CDATA[
<p>The entire ecosystem of CC is designed to facilitate burning tokens. You have to ask the LLM to write a script for the app to tell you which folder you're working in and which branch you're on. There are commands that just diagnose your Claude Code setup and try to "optimize" it. Adding skills or plugins bloats the context window. Developing plans means that you work through questions before you get to it in the code, but that matters way more for human programmers than LLMs, so it's probably just a waste of tokens.</p>
]]></description><pubDate>Fri, 07 Aug 2026 14:33:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=49211081</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49211081</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49211081</guid></item><item><title><![CDATA[New comment by km144 in "Qwen3.8 Max now ranked as the best overall model by agentic index"]]></title><description><![CDATA[
<p>What output style have you found to actually fix Opus 5's grating prose then? I find it leans hard into its preferred grammatical structures and rote sayings no matter what I include in the output style.</p>
]]></description><pubDate>Fri, 07 Aug 2026 14:30:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=49211025</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49211025</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49211025</guid></item><item><title><![CDATA[New comment by km144 in "Not hiring junior engineers won't solve the problem you think you have"]]></title><description><![CDATA[
<p>> because it's really hard to learn anything by skimming code that's being pooped out by Claude.<p>I think this is pretty dubious. It's true that you have to do more proactive learning with an agentic workflow, but it doesn't mean you can't learn stuff. Software engineers forged in the age of AI are not learning the same things their predecessors did or thinking about code in the same way, but the more curious people will still be better at the job, which is I think how it's always been. It seems a little unclear how much "AI brain rot" will be a force unto itself that neuters even the smartest of the juniors, but I think it's still an open question.</p>
]]></description><pubDate>Wed, 05 Aug 2026 19:37:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=49187850</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49187850</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49187850</guid></item><item><title><![CDATA[New comment by km144 in "Claude: Elevated errors across all models – Resolved"]]></title><description><![CDATA[
<p>You're right—and it's worth calling out explicitly, because that completely changes the approach here.</p>
]]></description><pubDate>Wed, 29 Jul 2026 20:21:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=49102527</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49102527</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49102527</guid></item><item><title><![CDATA[New comment by km144 in "Claude Opus 5"]]></title><description><![CDATA[
<p>Here is one data point for cost:<p><a href="https://artificialanalysis.ai/models?cost=intelligence-vs-cost-per-task&model-filters=large-models%2Cproprietary%2Creasoning-models#cost-tabs" rel="nofollow">https://artificialanalysis.ai/models?cost=intelligence-vs-co...</a><p>Here is another data point for output token efficiency:<p><a href="https://artificialanalysis.ai/models?cost=intelligence-vs-cost-per-task&model-filters=large-models%2Cproprietary%2Creasoning-models&intelligence-index-token-use=intelligence-vs-output-tokens-per-task#intelligence-index-token-use-tabs" rel="nofollow">https://artificialanalysis.ai/models?cost=intelligence-vs-co...</a></p>
]]></description><pubDate>Fri, 24 Jul 2026 18:04:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=49039456</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49039456</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49039456</guid></item><item><title><![CDATA[New comment by km144 in "Claude Opus 5"]]></title><description><![CDATA[
<p>I agree, very odd they did not comment on any theories for the degradation here. Dip and then rebound at max effort is pretty interesting too. Overthinking is bad, but you can overthink so much it starts to be better again?</p>
]]></description><pubDate>Fri, 24 Jul 2026 17:59:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=49039400</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=49039400</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49039400</guid></item><item><title><![CDATA[New comment by km144 in "Kernel anti-cheat is an overreach"]]></title><description><![CDATA[
<p>> So I would rather share a match with the occasional cheater than run un-auditable ring-0 software on the same machine I use for anything private.<p>The article makes an argument that anti-cheat is not worth the trade-off, yet the author admits they are a non-gamer. Then they go on to present one example of anti-cheat that tells us all we need to know about actual gamers' preferences—FACEIT. For those who don't know, FACEIT is a third-party matchmaking service, primarily for CS2. People <i>choose</i> to go through the hoops of using third-party service that installs kernel-level anti-cheat on their computer because it helps to keep cheaters out of their games. This seems like pretty strong evidence that the author's argument is not a good representation of gamers' thoughts on this. I don't know what the actual solution is. I suspect if Valve made their own kernel-level anti-cheat people might trust it more, but it's still the same problem.</p>
]]></description><pubDate>Tue, 07 Jul 2026 12:05:05 +0000</pubDate><link>https://news.ycombinator.com/item?id=48816579</link><dc:creator>km144</dc:creator><comments>https://news.ycombinator.com/item?id=48816579</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48816579</guid></item></channel></rss>