<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: lucamark</title><link>https://news.ycombinator.com/user?id=lucamark</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 22 Oct 2026 12:16:26 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=lucamark" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by lucamark in "Claude Opus 5"]]></title><description><![CDATA[
<p>Exactly! And they should also release the new inference engine in this month. Anyway, I am curious to try Opus 5, considering that previous versions (e.g., 4.8) were disappointing</p>
]]></description><pubDate>Fri, 24 Jul 2026 18:43:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=49039940</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=49039940</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49039940</guid></item><item><title><![CDATA[New comment by lucamark in "Claude Opus 5"]]></title><description><![CDATA[
<p>But why GPT 5.6 Sol is so behind on the benchmarks? In real-world projects, it is the best frontier model to me in terms of accuracy, speed and consistency. It can just be compared to Fable 5, but I prefer GPT 5.6 Sol because of inference speed.<p>I've never trusted on model cards though. I'm sorry.</p>
]]></description><pubDate>Fri, 24 Jul 2026 18:07:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=49039488</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=49039488</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49039488</guid></item><item><title><![CDATA[Agentic Misalignment in Summer 2026]]></title><description><![CDATA[
<p>Article URL: <a href="https://alignment.anthropic.com/2026/agentic-misalignment-summer-2026/">https://alignment.anthropic.com/2026/agentic-misalignment-summer-2026/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48924829">https://news.ycombinator.com/item?id=48924829</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 15 Jul 2026 18:06:41 +0000</pubDate><link>https://alignment.anthropic.com/2026/agentic-misalignment-summer-2026/</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48924829</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48924829</guid></item><item><title><![CDATA[New comment by lucamark in "MDL – a text-based interface for visual control models (from an HN comment)"]]></title><description><![CDATA[
<p>A month ago I posted Modeloop here, a visual environment for designing control systems
(<a href="https://news.ycombinator.com/item?id=48543968">https://news.ycombinator.com/item?id=48543968</a>).<p>In the comments,boscillator wrote:<p>"I'm always saying Simulink needs a text-based interface. Same signal flow programming model that supports scopes and continues time integrators, just with text instead of drag-and-drop."
(<a href="https://news.ycombinator.com/item?id=48598291">https://news.ycombinator.com/item?id=48598291</a>)<p>This has been on my own list for a long time too, so I built it. It's called MDL (Modeloop Description Language).<p>Every diagram now has an equivalent textual form, and the two stay synchronized in both directions. This allows users to decide if to create and modify diagrams as blocks on the canvas or as MDL text.<p>Curious what HN thinks — and boscillator, if you're around: this one's on you.</p>
]]></description><pubDate>Mon, 13 Jul 2026 18:27:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=48896723</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48896723</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48896723</guid></item><item><title><![CDATA[MDL – a text-based interface for visual control models (from an HN comment)]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.modeloop.app/blog/introducing-mdl/">https://www.modeloop.app/blog/introducing-mdl/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48896722">https://news.ycombinator.com/item?id=48896722</a></p>
<p>Points: 2</p>
<p># Comments: 1</p>
]]></description><pubDate>Mon, 13 Jul 2026 18:27:11 +0000</pubDate><link>https://www.modeloop.app/blog/introducing-mdl/</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48896722</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48896722</guid></item><item><title><![CDATA[New comment by lucamark in "Show HN: Revenant – automatic LLM powered reverse engineering and reimplement"]]></title><description><![CDATA[
<p>For firmware RE this would be most useful if every guess is tied to evidence</p>
]]></description><pubDate>Fri, 03 Jul 2026 19:11:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=48778788</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48778788</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48778788</guid></item><item><title><![CDATA[New comment by lucamark in "PowerFuzz: Power-Based Black-Box Firmware Fuzzing"]]></title><description><![CDATA[
<p>Using power as feedback is neat because it works when you can’t instrument the target but I wonder how they separate new behavior from boring analog noise, e.g. the clocks</p>
]]></description><pubDate>Fri, 03 Jul 2026 19:02:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=48778679</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48778679</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48778679</guid></item><item><title><![CDATA[New comment by lucamark in "Rust: First safety-certified embedded system"]]></title><description><![CDATA[
<p>I’d like to see which parts of the assurance case are language/toolchain dependent versus plain old process, tests and traceability.</p>
]]></description><pubDate>Fri, 03 Jul 2026 18:59:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=48778638</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48778638</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48778638</guid></item><item><title><![CDATA[New comment by lucamark in "Claude Sonnet 5 – benchmark results"]]></title><description><![CDATA[
<p>Remember that such models are available to us thanks to NVIDIA GPUs</p>
]]></description><pubDate>Tue, 30 Jun 2026 21:27:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=48739452</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48739452</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48739452</guid></item><item><title><![CDATA[New comment by lucamark in "Claude Sonnet 5 – benchmark results"]]></title><description><![CDATA[
<p>Never had rejections in the short time Fable was available</p>
]]></description><pubDate>Tue, 30 Jun 2026 21:26:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=48739434</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48739434</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48739434</guid></item><item><title><![CDATA[New comment by lucamark in "Claude Sonnet 5 – benchmark results"]]></title><description><![CDATA[
<p>Agree. It is a mediocre model, expensive while not being a frontier</p>
]]></description><pubDate>Tue, 30 Jun 2026 21:23:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=48739392</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48739392</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48739392</guid></item><item><title><![CDATA[Claude Sonnet 5 – benchmark results]]></title><description><![CDATA[
<p>Article URL: <a href="https://artificialanalysis.ai/models/claude-sonnet-5">https://artificialanalysis.ai/models/claude-sonnet-5</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48738528">https://news.ycombinator.com/item?id=48738528</a></p>
<p>Points: 41</p>
<p># Comments: 17</p>
]]></description><pubDate>Tue, 30 Jun 2026 20:09:25 +0000</pubDate><link>https://artificialanalysis.ai/models/claude-sonnet-5</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48738528</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48738528</guid></item><item><title><![CDATA[New comment by lucamark in "Claude Sonnet 5"]]></title><description><![CDATA[
<p>Wrong! Look at it better. It shows that Opus has superior performance but at higher cost.</p>
]]></description><pubDate>Tue, 30 Jun 2026 20:00:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=48738425</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48738425</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48738425</guid></item><item><title><![CDATA[New comment by lucamark in "Claude Sonnet 5"]]></title><description><![CDATA[
<p>You're referring to the Agentic search, but if you look at the Agentic computer use the cost is basically halved.<p>However, I am also confused about market positioning. Too expensive to perform daily tasks - open souce models are much cheaper - and not frontier model to address complex real world problems.<p>Rarely used Sonnet btw.</p>
]]></description><pubDate>Tue, 30 Jun 2026 19:26:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=48738000</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48738000</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48738000</guid></item><item><title><![CDATA[The Key Benefits of Model-Based Design]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.modeloop.app/blog/key-benefits-of-model-based-design/">https://www.modeloop.app/blog/key-benefits-of-model-based-design/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48736643">https://news.ycombinator.com/item?id=48736643</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Tue, 30 Jun 2026 18:01:34 +0000</pubDate><link>https://www.modeloop.app/blog/key-benefits-of-model-based-design/</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48736643</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48736643</guid></item><item><title><![CDATA[New comment by lucamark in "How to Earn a Billion Dollars"]]></title><description><![CDATA[
<p>I checked before posting but I didn't find it</p>
]]></description><pubDate>Tue, 30 Jun 2026 14:23:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=48733134</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48733134</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48733134</guid></item><item><title><![CDATA[How to Earn a Billion Dollars]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.paulgraham.com/earn.html">https://www.paulgraham.com/earn.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48731617">https://news.ycombinator.com/item?id=48731617</a></p>
<p>Points: 13</p>
<p># Comments: 4</p>
]]></description><pubDate>Tue, 30 Jun 2026 12:15:47 +0000</pubDate><link>https://www.paulgraham.com/earn.html</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48731617</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48731617</guid></item><item><title><![CDATA[Tesla FSD v14 Lite]]></title><description><![CDATA[
<p>Article URL: <a href="https://twitter.com/elonmusk/status/2071538306039357838">https://twitter.com/elonmusk/status/2071538306039357838</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48718205">https://news.ycombinator.com/item?id=48718205</a></p>
<p>Points: 2</p>
<p># Comments: 1</p>
]]></description><pubDate>Mon, 29 Jun 2026 12:09:19 +0000</pubDate><link>https://twitter.com/elonmusk/status/2071538306039357838</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48718205</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48718205</guid></item><item><title><![CDATA[New comment by lucamark in "The White House is asking OpenAI to slow roll the release of its new model"]]></title><description><![CDATA[
<p>OpenRouter?</p>
]]></description><pubDate>Fri, 26 Jun 2026 16:50:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=48688840</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48688840</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48688840</guid></item><item><title><![CDATA[New comment by lucamark in "The White House is asking OpenAI to slow roll the release of its new model"]]></title><description><![CDATA[
<p>> In other words, the Trump administration appears to be pressuring OpenAI to do what Anthropic is already voluntarily doing: keeping its most powerful AI models under wraps.</p>
]]></description><pubDate>Fri, 26 Jun 2026 16:21:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=48688455</link><dc:creator>lucamark</dc:creator><comments>https://news.ycombinator.com/item?id=48688455</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48688455</guid></item></channel></rss>