<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: arilotter</title><link>https://news.ycombinator.com/user?id=arilotter</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 13 Jun 2026 07:57:29 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=arilotter" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by arilotter in "Rust in the kernel is no longer experimental"]]></title><description><![CDATA[
<p>This title is moderately clickbait-y and comes with a subtle implication that Rust might be getting removed from the kernel. IMO it should be changed to "Rust in the kernel is no longer experimental"</p>
]]></description><pubDate>Wed, 10 Dec 2025 03:47:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=46213850</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=46213850</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46213850</guid></item><item><title><![CDATA[New comment by arilotter in "Democratizing AI: The Psyche Network Architecture"]]></title><description><![CDATA[
<p>Hi - Psyche developer here. Happy to answer any/all questions about the project :)</p>
]]></description><pubDate>Wed, 14 May 2025 21:51:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=43989583</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=43989583</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43989583</guid></item><item><title><![CDATA[New comment by arilotter in "Peer-to-peer file transfers in the browser"]]></title><description><![CDATA[
<p>iroh's stuff is great but their local peer discovery can't work in a browser, since it uses an mdns-like protocol to do it</p>
]]></description><pubDate>Wed, 12 Mar 2025 13:46:23 +0000</pubDate><link>https://news.ycombinator.com/item?id=43343211</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=43343211</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43343211</guid></item><item><title><![CDATA[New comment by arilotter in "Iroh – P2P that just works"]]></title><description><![CDATA[
<p>big ++ for iroh - using it for a project at work, it's evolving really nicely, has reliable holepunching, and the team is super responsive and happy to dig thru gigs of logs to find a bug :)</p>
]]></description><pubDate>Tue, 11 Mar 2025 00:56:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=43328041</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=43328041</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43328041</guid></item><item><title><![CDATA[New comment by arilotter in "Supreme Court upholds TikTok ban, but Trump might offer lifeline"]]></title><description><![CDATA[
<p>It's unlikely you can. They're generally only willing to set up corporate deals to sell data in massive bulk. You could buy a domain name, set up a decently real looking website with a corpo looking email, then go to any broker like <a href="https://www.acxiom.com/customer-data/" rel="nofollow">https://www.acxiom.com/customer-data/</a> (not affiliated, first one i found on google) and do the whole corporate dance of signing a contract to get what you want.</p>
]]></description><pubDate>Fri, 17 Jan 2025 21:51:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=42743644</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=42743644</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42743644</guid></item><item><title><![CDATA[New comment by arilotter in "Fluid Simulation Pendant"]]></title><description><![CDATA[
<p>I learned recently that Charlieplexing is kind of like the discrete version of the Radon Transform: <a href="https://en.wikipedia.org/wiki/Radon_transform" rel="nofollow">https://en.wikipedia.org/wiki/Radon_transform</a></p>
]]></description><pubDate>Tue, 14 Jan 2025 00:21:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=42691594</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=42691594</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42691594</guid></item><item><title><![CDATA[New comment by arilotter in "Pre-training a 15B parameter language model over the internet"]]></title><description><![CDATA[
<p>you could absolutely use DisTrO for federated learning. The DeMo optimizer on its own doesn't solve the adverserial aspects of training on local-only data, nor does it solve tensor parallelism across devices, so you're still limited to only what fits on your local GPUs, but it <i>does</i> enable distributed <i>data parallelism</i> over the internet at a bandwidth orders of magnitude lower than before.</p>
]]></description><pubDate>Mon, 02 Dec 2024 19:59:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=42299762</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=42299762</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42299762</guid></item><item><title><![CDATA[New comment by arilotter in "Pre-training a 15B parameter language model over the internet"]]></title><description><![CDATA[
<p>This specific model is only trained on 100 billion tokens, so it's not SOTA by any means, but we've got designs on larger training runs later :)</p>
]]></description><pubDate>Mon, 02 Dec 2024 19:57:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=42299741</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=42299741</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42299741</guid></item><item><title><![CDATA[New comment by arilotter in "Pre-training a 15B parameter language model over the internet"]]></title><description><![CDATA[
<p>> Are you planning to eventually do a SETI@home style thing where anyone can join?<p>That's one of the goals of the stuff we're working on right now, I personally hope we can make it work!<p>> What was Durk Kingma's involvement in DisTrO?<p>As a co-author on the paper, we brought him in to bounce some ideas off of & have him validate DeMo's design and implementation to ensure we weren't hallucinating these results.</p>
]]></description><pubDate>Mon, 02 Dec 2024 19:56:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=42299738</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=42299738</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42299738</guid></item><item><title><![CDATA[New comment by arilotter in "Pre-training a 15B parameter language model over the internet"]]></title><description><![CDATA[
<p>Yo! I'm on the team at Nous that built this tech, happy to answer any questions :)</p>
]]></description><pubDate>Mon, 02 Dec 2024 18:06:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=42298646</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=42298646</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42298646</guid></item><item><title><![CDATA[New comment by arilotter in "DisTrO – a family of low latency distributed optimizers"]]></title><description><![CDATA[
<p>That's not something that DisTrO solves, but there's plenty of research in that area! See <a href="https://arxiv.org/abs/2301.11913" rel="nofollow">https://arxiv.org/abs/2301.11913</a> , <a href="https://arxiv.org/abs/2206.01288" rel="nofollow">https://arxiv.org/abs/2206.01288</a> , <a href="https://arxiv.org/abs/2304.11277" rel="nofollow">https://arxiv.org/abs/2304.11277</a> etc :)</p>
]]></description><pubDate>Wed, 28 Aug 2024 18:19:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=41382521</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41382521</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41382521</guid></item><item><title><![CDATA[New comment by arilotter in "DisTrO – a family of low latency distributed optimizers"]]></title><description><![CDATA[
<p>we're excited to drop the paper & code and let you run the code for yourself to see otherwise !! coming in the next couple months :)</p>
]]></description><pubDate>Wed, 28 Aug 2024 01:45:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=41375197</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41375197</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41375197</guid></item><item><title><![CDATA[New comment by arilotter in "DisTrO – a family of low latency distributed optimizers"]]></title><description><![CDATA[
<p>We're excited about the potential and want to find other folks also excited about it that are interested in working for/with us to build things on the foundations of DisTrO!
Plus also it's so cool and mind boggling to us that we wanted to share the hype a little bit, it was hard not being able to tell anyone we were working on it</p>
]]></description><pubDate>Tue, 27 Aug 2024 21:19:08 +0000</pubDate><link>https://news.ycombinator.com/item?id=41373183</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41373183</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41373183</guid></item><item><title><![CDATA[New comment by arilotter in "DisTrO – a family of low latency distributed optimizers"]]></title><description><![CDATA[
<p>Paper & code in the next couple months. We're workin on em :)</p>
]]></description><pubDate>Tue, 27 Aug 2024 21:18:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=41373176</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41373176</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41373176</guid></item><item><title><![CDATA[New comment by arilotter in "DisTrO – a family of low latency distributed optimizers"]]></title><description><![CDATA[
<p>We're pretty confident it's not a fluke, and paper + code are the next step, within a couple months.
It's not "synchronize every step", but it's "do something every step".<p>We double and triple and quadruple checked our results, to make sure that we are in fact getting results like this while only doing our thing every step, and it really keeps holding up.<p>Don't trust our word for it, though, you'll see when the paper comes out :)</p>
]]></description><pubDate>Tue, 27 Aug 2024 20:38:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=41372718</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41372718</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41372718</guid></item><item><title><![CDATA[New comment by arilotter in "Show HN: Nous – Open-Source Agent Framework with Autonomous, SWE Agents, WebUI"]]></title><description><![CDATA[
<p>I can confirm this is not a projected related to Nous Research in any way, just an unfortunate naming collision</p>
]]></description><pubDate>Fri, 09 Aug 2024 17:37:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=41203842</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41203842</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41203842</guid></item><item><title><![CDATA[New comment by arilotter in "macOS in QEMU in Docker"]]></title><description><![CDATA[
<p>i remember making this joke a decade ago on the tonymacx86 forums, lots of people did it for the lulz</p>
]]></description><pubDate>Wed, 31 Jul 2024 18:05:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=41121599</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41121599</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41121599</guid></item><item><title><![CDATA[New comment by arilotter in "macOS in QEMU in Docker"]]></title><description><![CDATA[
<p>you can! you can also run it on a real iOS device using this tech. it's explicitly documented in the repo :)</p>
]]></description><pubDate>Wed, 31 Jul 2024 18:04:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=41121591</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41121591</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41121591</guid></item><item><title><![CDATA[New comment by arilotter in "Meta has run hundreds of ads for cocaine, opioids and other drugs"]]></title><description><![CDATA[
<p>Anecdotally, I get Instagram ads for "buy mdma online" all day, every day.</p>
]]></description><pubDate>Wed, 31 Jul 2024 18:03:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=41121578</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=41121578</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41121578</guid></item><item><title><![CDATA[New comment by arilotter in "World_sim: LLM prompted to act as a sentient CLI universe simulator"]]></title><description><![CDATA[
<p>(I work at Nous) it's Anthropic's Claude 3 Opus!
Working around rejections is always tricky, and you gotta juggle getting responses to interesting queries with not breaking Anthropic's ToS</p>
]]></description><pubDate>Sat, 06 Apr 2024 05:00:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=39950112</link><dc:creator>arilotter</dc:creator><comments>https://news.ycombinator.com/item?id=39950112</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39950112</guid></item></channel></rss>