<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: jpardoe</title><link>https://news.ycombinator.com/user?id=jpardoe</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 18 Nov 2026 03:44:57 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=jpardoe" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by jpardoe in "GPT‑6.2 Astra and Sol"]]></title><description><![CDATA[
<p>This is why I use Opus as planner/orchestrator and Sol/Terra(max) implementers. Opus is just way better at handling long running tasks/projects and the OpenAI models are better at smaller, well defined tasks. Claude alone is fine but I max out my $200 sub too often so the extra work comes from OpenAI.</p>
]]></description><pubDate>Wed, 18 Nov 2026 03:38:04 +0000</pubDate><link>https://news.ycombinator.com/item?id=50541557</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50541557</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50541557</guid></item><item><title><![CDATA[New comment by jpardoe in "GPT‑6.2 Astra and Sol"]]></title><description><![CDATA[
<p>I’m sorry but this is crazy, I think you’ve forgotten just how bad 5 was. It’s not perfect by any means but it’s a clear improvement.</p>
]]></description><pubDate>Tue, 17 Nov 2026 19:05:32 +0000</pubDate><link>https://news.ycombinator.com/item?id=50536521</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50536521</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50536521</guid></item><item><title><![CDATA[New comment by jpardoe in "GPT‑6.2 Astra and Sol"]]></title><description><![CDATA[
<p>That’s exactly what I do. It’s more complex than this, but vaguely: Opus medium as orchestrator and reviewer, Sol(high/medium)/Terra(max) as implementer. My set up looks at my next GPT reset date and dynamically enters FAST mode if I’m under achieving on projected usage.</p>
]]></description><pubDate>Tue, 17 Nov 2026 18:52:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=50536326</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50536326</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50536326</guid></item><item><title><![CDATA[New comment by jpardoe in "Tell HN: HN is being scraped for emails to scam"]]></title><description><![CDATA[
<p>Of course.</p>
]]></description><pubDate>Mon, 16 Nov 2026 08:47:09 +0000</pubDate><link>https://news.ycombinator.com/item?id=50515440</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50515440</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50515440</guid></item><item><title><![CDATA[New comment by jpardoe in "Serving GLM-5.3 at 607.8 tok/s across two Nvidia DGX stations with vLLM"]]></title><description><![CDATA[
<p>I don't know what you mean.</p>
]]></description><pubDate>Mon, 16 Nov 2026 05:40:57 +0000</pubDate><link>https://news.ycombinator.com/item?id=50514423</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50514423</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50514423</guid></item><item><title><![CDATA[New comment by jpardoe in "Serving GLM-5.3 at 607.8 tok/s across two Nvidia DGX stations with vLLM"]]></title><description><![CDATA[
<p>No. You can’t compete with large data centres. You will end up paying more in energy and hardware than what you charge for the compute.</p>
]]></description><pubDate>Sun, 15 Nov 2026 08:24:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=50505092</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50505092</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50505092</guid></item><item><title><![CDATA[New comment by jpardoe in "AI is destroying the trust between writers and readers"]]></title><description><![CDATA[
<p>I’m hopeful but not optimistic that AI will ruin the internet so badly that there will be a reversion to the real world as people crave authenticity.</p>
]]></description><pubDate>Sat, 14 Nov 2026 12:30:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=50496513</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50496513</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50496513</guid></item><item><title><![CDATA[New comment by jpardoe in "Researchers detail over 1,500 verified cases of children with past-life memories"]]></title><description><![CDATA[
<p>This is very hard to believe, and I’m not saying you’re lying. He was reading like an 8 year old at 3 years old but only “did well at school”? I mean he would have been crushing the other kids with that level of advancement, surely?</p>
]]></description><pubDate>Fri, 13 Nov 2026 16:06:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=50485549</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50485549</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50485549</guid></item><item><title><![CDATA[New comment by jpardoe in "Researchers detail over 1,500 verified cases of children with past-life memories"]]></title><description><![CDATA[
<p>You’re telling me your 3 year old could read the words “Nintendo advanced reviews”?</p>
]]></description><pubDate>Fri, 13 Nov 2026 12:40:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=50482687</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50482687</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50482687</guid></item><item><title><![CDATA[New comment by jpardoe in "Researchers detail over 1,500 verified cases of children with past-life memories"]]></title><description><![CDATA[
<p>Dripping with irony.</p>
]]></description><pubDate>Fri, 13 Nov 2026 12:37:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=50482659</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50482659</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50482659</guid></item><item><title><![CDATA[New comment by jpardoe in "FBI says US has seized $1.2B worth of cryptocurrency stolen from Binance"]]></title><description><![CDATA[
<p>It’s likely the stable coins. Those can be frozen by the stable coin operators.</p>
]]></description><pubDate>Fri, 13 Nov 2026 08:23:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=50480103</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50480103</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50480103</guid></item><item><title><![CDATA[New comment by jpardoe in "Anthropic IPO"]]></title><description><![CDATA[
<p>The frontier is addictive.</p>
]]></description><pubDate>Thu, 12 Nov 2026 13:23:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50466723</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50466723</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50466723</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>That makes no sense.</p>
]]></description><pubDate>Wed, 11 Nov 2026 12:03:39 +0000</pubDate><link>https://news.ycombinator.com/item?id=50452433</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50452433</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50452433</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>I’ll use it if there’s a complex bug I need solving quickly, or to weigh in on architectural decisions across large multi-plan projects. It’s a tool to reach for but makes no sense as a daily driver.</p>
]]></description><pubDate>Tue, 10 Nov 2026 06:53:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=50436929</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50436929</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50436929</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>Distilled, weaker versions though I think?</p>
]]></description><pubDate>Tue, 10 Nov 2026 06:22:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=50436756</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50436756</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50436756</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>You will get more done, at higher quality and quicker with an Anthropic sub compared to using DeepSeek. Unless you’re forced to use API billing the open weight models make no sense right now.</p>
]]></description><pubDate>Tue, 10 Nov 2026 06:21:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=50436750</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50436750</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50436750</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>I really hope so.</p>
]]></description><pubDate>Tue, 10 Nov 2026 06:17:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=50436719</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50436719</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50436719</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>Exactly the same experience on both subscriptions. Claude is noticeably better value. If things were frozen in time and I could only choose one it would be Claude, no question.</p>
]]></description><pubDate>Tue, 10 Nov 2026 06:16:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=50436714</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50436714</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50436714</guid></item><item><title><![CDATA[New comment by jpardoe in "Claude Fable 5.2 scores 65 on Artificial Analysis, highest so far"]]></title><description><![CDATA[
<p>All I know is recently I’ve been switching in and out of Codex and Claude with the same workflow and my Codex usage gets drained much quicker. Just my experience.</p>
]]></description><pubDate>Tue, 10 Nov 2026 06:13:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=50436696</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50436696</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50436696</guid></item><item><title><![CDATA[New comment by jpardoe in "Goodbye Claude, hello Pi"]]></title><description><![CDATA[
<p>I'm not saying you don't get good value out of your $100, I'm sure you're doing great, but I'm also absolutely certain if you followed the exact same workflow with an Anthropic subscription you would get better value.</p>
]]></description><pubDate>Mon, 09 Nov 2026 10:54:41 +0000</pubDate><link>https://news.ycombinator.com/item?id=50423883</link><dc:creator>jpardoe</dc:creator><comments>https://news.ycombinator.com/item?id=50423883</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50423883</guid></item></channel></rss>