<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: DenisM</title><link>https://news.ycombinator.com/user?id=DenisM</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 25 Jul 2026 23:53:02 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=DenisM" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by DenisM in "AI Companies Are Trying to Hide a Staggering Amount of Debt"]]></title><description><![CDATA[
<p>Not quite a Bear Sterns scale event.</p>
]]></description><pubDate>Fri, 24 Jul 2026 17:04:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=49038547</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49038547</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49038547</guid></item><item><title><![CDATA[New comment by DenisM in "Why is programming so captivating, yet so agonizing?"]]></title><description><![CDATA[
<p>How do you know what you built is good? That was the pivot for me - I defined it for myself as it’s something that makes a positive contribution to my users.</p>
]]></description><pubDate>Fri, 24 Jul 2026 15:40:52 +0000</pubDate><link>https://news.ycombinator.com/item?id=49037286</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49037286</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49037286</guid></item><item><title><![CDATA[New comment by DenisM in "Government orders GitHub to remove Bluetooth-based chat app Bitchat: Jack Dorsey"]]></title><description><![CDATA[
<p>For clarity: Indian government.</p>
]]></description><pubDate>Fri, 24 Jul 2026 15:11:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=49036855</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49036855</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49036855</guid></item><item><title><![CDATA[New comment by DenisM in "How much energy do data centers and artificial intelligence use?"]]></title><description><![CDATA[
<p>Remarkably, it took non-AI workloads 25 years to grow from zero to 3%, and it took AI only ~ 2.5 years to get to 2%.<p>This puts a number to my intuitive feeling that things move an order of magnitude faster this time.</p>
]]></description><pubDate>Thu, 23 Jul 2026 17:02:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=49024822</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49024822</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49024822</guid></item><item><title><![CDATA[New comment by DenisM in "How much energy do data centers and artificial intelligence use?"]]></title><description><![CDATA[
<p>Where did the tokens go?</p>
]]></description><pubDate>Thu, 23 Jul 2026 16:35:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=49024404</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49024404</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49024404</guid></item><item><title><![CDATA[New comment by DenisM in "How much energy do data centers and artificial intelligence use?"]]></title><description><![CDATA[
<p>> So what about the other 3%? That's not "massive"? Or it gets a pass because it's not AI?<p>How did you read that in my comment?</p>
]]></description><pubDate>Thu, 23 Jul 2026 16:32:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=49024350</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49024350</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49024350</guid></item><item><title><![CDATA[New comment by DenisM in "How much energy do data centers and artificial intelligence use?"]]></title><description><![CDATA[
<p>Select quote:<p>* 5% of electricity in the United States is used to power data centers. For AI-focused ones specifically, it’s probably around 2%.*<p>I think it’s massive.</p>
]]></description><pubDate>Thu, 23 Jul 2026 15:46:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=49023498</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49023498</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49023498</guid></item><item><title><![CDATA[New comment by DenisM in "AI Companies Are Trying to Hide a Staggering Amount of Debt"]]></title><description><![CDATA[
<p>This media frenzy can go on forever tho.<p>A Bear Sterns moment would be more solid. Oracle might ge the first to collapse if things go south, so we might be fine until then(?).</p>
]]></description><pubDate>Thu, 23 Jul 2026 15:34:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=49023272</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=49023272</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49023272</guid></item><item><title><![CDATA[New comment by DenisM in "Harness Engineering"]]></title><description><![CDATA[
<p>In my little ai bubble the original blog post was very influential. It does sound immodest, but could also be true.</p>
]]></description><pubDate>Sun, 19 Jul 2026 17:50:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=48970192</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48970192</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48970192</guid></item><item><title><![CDATA[New comment by DenisM in "Texas wins court order to suspend domain name for violating age-verification law"]]></title><description><![CDATA[
<p>Register two domains in different jurisdictions, with neither serving its own jurisdiction, and a redirect for stray users.</p>
]]></description><pubDate>Sat, 18 Jul 2026 01:17:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=48954171</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48954171</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48954171</guid></item><item><title><![CDATA[New comment by DenisM in "Show HN: Tarit – a hypervisor which is 2x faster than firecracker"]]></title><description><![CDATA[
<p>Not clear where 2x claim in the headline came from, I don’t see it in the linked site.</p>
]]></description><pubDate>Fri, 17 Jul 2026 22:35:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=48952937</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48952937</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48952937</guid></item><item><title><![CDATA[New comment by DenisM in "The human-in-the-loop is tired"]]></title><description><![CDATA[
<p>> the runaway inference truck.<p>Nicely done, thank you.</p>
]]></description><pubDate>Fri, 17 Jul 2026 05:37:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=48943668</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48943668</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48943668</guid></item><item><title><![CDATA[New comment by DenisM in "Towards a harness that can do anything"]]></title><description><![CDATA[
<p>No sarcasm. In that thread we discussed recursive, interleaved  application of code and prompt. No morphing, however.<p>Thanks for dynamic workflow pointer. I don’t know if I like JavaScript for workflow definitions tho. IMHO sshwarts has the right idea on severely constrained workflow definition language. I plan to look closer into ADK2 workflows as well.<p>I’m not saying anyone did what you are doing, I’m saying multiple pieces are converging on that, at least in my imagination.</p>
]]></description><pubDate>Thu, 16 Jul 2026 01:20:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=48929345</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48929345</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48929345</guid></item><item><title><![CDATA[New comment by DenisM in "Towards a harness that can do anything"]]></title><description><![CDATA[
<p>The idea of morphing prompts to code and back is revolutionary.<p>You may want to read earlier discussions <a href="https://news.ycombinator.com/item?id=48881112">https://news.ycombinator.com/item?id=48881112</a>
And <a href="https://news.ycombinator.com/item?id=48051562">https://news.ycombinator.com/item?id=48051562</a><p>I</p>
]]></description><pubDate>Wed, 15 Jul 2026 20:57:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=48926916</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48926916</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48926916</guid></item><item><title><![CDATA[New comment by DenisM in "Show HN: Skillscript – A declarative, sandboxed language for tool orchestration"]]></title><description><![CDATA[
<p>> I tested the language against cold agents<p>I was procrastinating on that, thanks!<p>> the recursive nature<p>It’s not clear from you answer if we’re on the same page. Your options are allow the worker agent run wild or constrain it with a rigid guidance. Skillscript can serve dual purpose - constrain the workflow to make it more observable and constrain elemental parts to make them more directional. Apologies if that makes no sense - the AI red shift (expanding universe) makes it very hard to communicate.</p>
]]></description><pubDate>Mon, 13 Jul 2026 16:49:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=48895364</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48895364</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48895364</guid></item><item><title><![CDATA[New comment by DenisM in "Show HN: Skillscript – A declarative, sandboxed language for tool orchestration"]]></title><description><![CDATA[
<p>1. How a this better than just using any other script language?<p>2. Inventing a new language complicates large models ability to generate such scrips compared to a well-known language. Did you find it to be a problem? How did you mitigate?<p>3. The AI is showing. :) I had the similar discussion with ChapGPT and some phrasing is near the same. Not a dig, just a funny observation.<p>4. Consider the recursive nature of the problem you’re solving - large model updates  workflow which you review each time, worker models generate plans and tool calls which you don’t review. A constrained language is useful in both cases to guide the model.<p>5. This Earlier discussion can provide useful background for why this is needed. You have probably seen it, but the readers will likely appreciate.   <a href="https://news.ycombinator.com/item?id=48051562">https://news.ycombinator.com/item?id=48051562</a><p>The problem is real, Thank you for taking a stab and sharing your findings.</p>
]]></description><pubDate>Sun, 12 Jul 2026 18:14:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=48883172</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48883172</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48883172</guid></item><item><title><![CDATA[New comment by DenisM in "GPT-5.6 Sol Ultra will be in Codex"]]></title><description><![CDATA[
<p>You seem to inhabit a hero/villaim dichotomy, and a narrative one at that. It’s not a place I wish to revisit.</p>
]]></description><pubDate>Thu, 09 Jul 2026 19:34:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=48851289</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48851289</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48851289</guid></item><item><title><![CDATA[New comment by DenisM in "Show HN: Docx-CLI: agents read/edit Word docs using 1/2 the time and tokens"]]></title><description><![CDATA[
<p>For A while I was expecting that MCP will dominate, but we seem to be going in the direction of CLI being more prevalent. Can’t wrap my mind around it.</p>
]]></description><pubDate>Wed, 08 Jul 2026 04:27:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=48827517</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48827517</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48827517</guid></item><item><title><![CDATA[New comment by DenisM in "Show HN: Docx-CLI: agents read/edit Word docs using 1/2 the time and tokens"]]></title><description><![CDATA[
<p>How would you compare that approach to spawning sub agents to operate high-token tools?</p>
]]></description><pubDate>Wed, 08 Jul 2026 03:36:58 +0000</pubDate><link>https://news.ycombinator.com/item?id=48827187</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48827187</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48827187</guid></item><item><title><![CDATA[New comment by DenisM in "GPT-5.6 Sol Ultra will be in Codex"]]></title><description><![CDATA[
<p>It stops being interesting, or even sad, after a while. People get stuck in all kinds of places, mentally. Some get unstuck eventually. It’s only sad if you have come to a counter factual belief that it could have gone better.<p>I went in the opposite direction - how far can I push myself to see multiple facets of a story? That is a wild ride, and it gets progressively more wild.</p>
]]></description><pubDate>Mon, 06 Jul 2026 04:42:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=48800711</link><dc:creator>DenisM</dc:creator><comments>https://news.ycombinator.com/item?id=48800711</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48800711</guid></item></channel></rss>