<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: mkaic</title><link>https://news.ycombinator.com/user?id=mkaic</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 25 Apr 2026 19:10:32 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=mkaic" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by mkaic in "Discord will require a face scan or ID for full access next month"]]></title><description><![CDATA[
<p>Just cancelled mine after reading this comment, I only really cared about the bigger file uploads and the HD screen-sharing anyways and I can live without those.<p>Now that I think of it, I bet I could host a decent instance of some open-source alternative in a public cloud for around the same cost as what I paid for Nitro ($100 a year)...</p>
]]></description><pubDate>Mon, 09 Feb 2026 20:54:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=46951117</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=46951117</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46951117</guid></item><item><title><![CDATA[New comment by mkaic in "Mistral OCR 3"]]></title><description><![CDATA[
<p>I suppose both of us watched the same Youtube video by Metta Beshay (i think that is his name?)</p>
]]></description><pubDate>Sat, 20 Dec 2025 06:37:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=46334082</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=46334082</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46334082</guid></item><item><title><![CDATA[New comment by mkaic in "VPN location claims don't match real traffic exits"]]></title><description><![CDATA[
<p>perhaps I shouldn't share my workaround, but I've found that Mullvad's Norway nodes <i>consistently</i> get past Reddit's IP-blocking :)</p>
]]></description><pubDate>Sun, 14 Dec 2025 01:12:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=46259934</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=46259934</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46259934</guid></item><item><title><![CDATA[New comment by mkaic in "Why more American seniors are getting high"]]></title><description><![CDATA[
<p><i>Heavy</i> and <i>chronic</i> usage of cannabis is <i>associated</i> with some of these things. I think "plenty of evidence that marijuana can cause {effects}" is somewhat overstating the consensus on the topic.<p>Like many things in life, cannabis can be enjoyed responsibly <i>or</i> irresponsibly. Irresponsible use is inadvisable and can absolutely ruin your life and the lives of others around you. I see no issue with responsible use, though. All things in moderation. Alcohol, social media, and caffeine all come to mind as examples of other drugs that can be largely safe and enjoyable if used responsibly, but which become dangerous/harmful when moderation is abandoned.<p>Signed,<p>a responsible/occasional cannabis user :)</p>
]]></description><pubDate>Fri, 12 Dec 2025 17:22:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=46246251</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=46246251</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46246251</guid></item><item><title><![CDATA[New comment by mkaic in "Google unkills JPEG XL?"]]></title><description><![CDATA[
<p>"It's... a regional dialect."<p>"What region?"<p>"Er, upstate New York."<p>"Really. Well, I'm from Utica and I've never heard anyone use the phrase '100M' to mean '100 thousand'"<p>"Oh, no, not in Utica. It's an Albany expression."</p>
]]></description><pubDate>Mon, 01 Dec 2025 19:01:31 +0000</pubDate><link>https://news.ycombinator.com/item?id=46111514</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=46111514</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46111514</guid></item><item><title><![CDATA[New comment by mkaic in "Blender 5.0"]]></title><description><![CDATA[
<p>I think some of the coolest changes in this release are on the nodes side of things — they added Closures (kinda like lambdas!), Bundles (tuples/structs, I guess?), and Repeat (loops!) (already was in Geo Nodes, now it's in Shader Nodes too).<p>Blender nodes have come a <i>long</i> way over the past decade and it's incredibly satisfying to see the care with which they have been developed. Blender's node editor is my personal favorite node editor I've ever used in any software, and I often find myself wishing other software adopted some of their UI and UX conventions.<p>Been a happy user since, oh, v2.75? And looking forward to being a user for many more releases to come.<p>Donate to Blender! [0]<p>[0] <a href="https://fund.blender.org" rel="nofollow">https://fund.blender.org</a></p>
]]></description><pubDate>Wed, 19 Nov 2025 06:14:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=45976406</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=45976406</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45976406</guid></item><item><title><![CDATA[New comment by mkaic in "By the Power of Grayscale"]]></title><description><![CDATA[
<p>IIIII HHHAAAAAVE THE POWERRRRR</p>
]]></description><pubDate>Tue, 04 Nov 2025 22:09:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=45816422</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=45816422</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45816422</guid></item><item><title><![CDATA[New comment by mkaic in "Ask HN: Who wants to be hired? (September 2025)"]]></title><description><![CDATA[
<p>Location: Boulder, Colorado<p>Remote: Open to Remote, Hybrid, or In-Person<p>Willing to relocate: Only for an exceptional offer.<p>Technologies: AI / Machine Learning, Docker, Python, PyTorch, Numpy, Pandas, Linux, TrueNAS, Pandas, SQL, Blender, Git / Github<p>Résumé/CV: <a href="https://docs.google.com/document/d/1io7PASX56d7eJTCXQnPxjqy3q7OLlTOZvzo5WuxgH2k/edit?usp=sharing" rel="nofollow">https://docs.google.com/document/d/1io7PASX56d7eJTCXQnPxjqy3...</a><p>Email: matthew.kai.christensen (at) gmail.com<p>Hello, my name is Kai and I'm a machine learning researcher with an interest in computer vision and novel neural network architecture design. I have 4 years of experience working on computer vision models for medical imagery, including publishing an echocardiography-focused finetune of CLIP in Nature Medicine last year. I'm self taught and enjoy work where I can be creative and learn new skills. I would love to be a part of the team that invents the Transformer-killer. I'm also open to more data-sciencey jobs if the work is interesting.<p>I also have 3D modeling, filmmaking/editing, and music-making skills and experience if they're relevant. I've been making movies and doing VFX ever since I was little.</p>
]]></description><pubDate>Thu, 04 Sep 2025 01:49:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=45122559</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=45122559</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45122559</guid></item><item><title><![CDATA[New comment by mkaic in "Making Minecraft Spherical"]]></title><description><![CDATA[
<p>My favorite Spherical Minecraft-like gamedev project is PlanetSmith [0], which uses hexagonal voxels (and a few pentagonal voxels). The devlogs are very well produced and I highly recommend checking them out.<p>[0] <a href="https://youtube.com/@incandescentgames" rel="nofollow">https://youtube.com/@incandescentgames</a></p>
]]></description><pubDate>Mon, 01 Sep 2025 17:27:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=45094717</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=45094717</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45094717</guid></item><item><title><![CDATA[New comment by mkaic in "A simple way to generate random points on a sphere"]]></title><description><![CDATA[
<p>My favorite way to generate random points on a n-dimensional sphere is to just sample n times from a Gaussian distribution to get a n-dimensional vector, and then normalizing that vector to the radius of the desired sphere.</p>
]]></description><pubDate>Sat, 23 Aug 2025 20:38:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=44998948</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44998948</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44998948</guid></item><item><title><![CDATA[Tape Speed Keyboard [video]]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.youtube.com/watch?v=adxlFwIGTfc">https://www.youtube.com/watch?v=adxlFwIGTfc</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=44770721">https://news.ycombinator.com/item?id=44770721</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 02 Aug 2025 19:42:09 +0000</pubDate><link>https://www.youtube.com/watch?v=adxlFwIGTfc</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44770721</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44770721</guid></item><item><title><![CDATA[New comment by mkaic in "Blender: Beyond Mouse and Keyboard"]]></title><description><![CDATA[
<p>Have you kept up with recent ML papers like MindEye, which have managed to reconstruct seen images using image generator models conditioned on fMRI signals?<p>Ever since that paper came out, I (someone who works in ML but have no neuroimaging expertise) have been really excited for the future of noninvasive BCU.<p>Would also be curious to know if you have any thoughts on the several start-ups working in parallel on optimally pumped magnetometers for portable MEG helmets.</p>
]]></description><pubDate>Mon, 28 Jul 2025 04:02:42 +0000</pubDate><link>https://news.ycombinator.com/item?id=44707126</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44707126</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44707126</guid></item><item><title><![CDATA[New comment by mkaic in "The Minecraft game score unexpectedly became big business for its composer"]]></title><description><![CDATA[
<p>I second this recommendation, I've been listening to the Wanderstop OST on shuffle for over a month now and find it delightful. The sister album "Wanderstop FM", which contains tracks that play diagetically on the in-game radio, is also excellent.</p>
]]></description><pubDate>Sun, 20 Jul 2025 17:07:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=44627118</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44627118</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44627118</guid></item><item><title><![CDATA[New comment by mkaic in "Being too ambitious is a clever form of self-sabotage"]]></title><description><![CDATA[
<p>I feel a bit shaken after reading this comment, to be honest. I don't think I've ever heard someone so perfectly describe such a major component of my life experience. It's like you read my mind.<p>I was a "gifted kid", now I'm a lonely adult living by herself constantly cycling between complacency, failure, panic, and productivity. Diagnosed ADHD, choose to stay unmedicated, sometimes the best employee in my office, usually one of the laziest and most disappointing employees in my office. Constantly daydreaming about how better circumstances would change things for the better even while knowing deep down I'd cause the exact same set of problems for myself all over again even if I got my Dream Job.<p>Spent my whole life being told I was exceptional, and, to be fair, I lived up to it as a kid. These days I'm so terrified of regressing to being "normal" that I sabotage myself at every turn.<p>Thank you for leaving this comment. I may bring up the concept with my therapist and see what she thinks of it.</p>
]]></description><pubDate>Sat, 05 Jul 2025 21:09:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=44475567</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44475567</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44475567</guid></item><item><title><![CDATA[New comment by mkaic in "Alternative Layout System"]]></title><description><![CDATA[
<p>I think in casual speech at this point (at least in my experience) the two are used interchangeably. In professional or legal settings I'm sure the distinction matters more, but I feel like OP's usage here felt pretty natural to me even though it's not technically correct.</p>
]]></description><pubDate>Fri, 27 Jun 2025 17:11:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=44398481</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44398481</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44398481</guid></item><item><title><![CDATA[New comment by mkaic in "Python can run Mojo now"]]></title><description><![CDATA[
<p>If they can manage to make good on their plans to open-source it, I'll breathe a tentative sigh of relief. I'm also rooting for them, but until they're open-source, I'm not willing to invest my own time into their ecosystem.</p>
]]></description><pubDate>Mon, 23 Jun 2025 03:48:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=44352306</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44352306</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44352306</guid></item><item><title><![CDATA[New comment by mkaic in "A South Korean grand master on the art of the perfect soy sauce"]]></title><description><![CDATA[
<p>Here's an unpaywalled PDF I found of the same: <a href="https://labs.la.utexas.edu/gilden/files/2016/04/theketchupconundrum.pdf" rel="nofollow">https://labs.la.utexas.edu/gilden/files/2016/04/theketchupco...</a></p>
]]></description><pubDate>Fri, 23 May 2025 04:10:27 +0000</pubDate><link>https://news.ycombinator.com/item?id=44069788</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=44069788</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44069788</guid></item><item><title><![CDATA[New comment by mkaic in "Flattening Rust’s learning curve"]]></title><description><![CDATA[
<p>I love the look and feel of Nim, but found it to be stuck in a weird chicken-and-egg situation where it didn't have enough of a following to have a Convenient Package For Everything, ultimately turning me off it. Of course I recognize that the only way a language <i>gets</i> a Convenient Package For Everything is if it gets <i>popular</i>, but still...</p>
]]></description><pubDate>Wed, 14 May 2025 19:04:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=43988114</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=43988114</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43988114</guid></item><item><title><![CDATA[How do you print a 3D photo? Gaussian Splats in resin [video]]]></title><description><![CDATA[

<p>Comments URL: <a href="https://news.ycombinator.com/item?id=43951043">https://news.ycombinator.com/item?id=43951043</a></p>
<p>Points: 1</p>
<p># Comments: 3</p>
]]></description><pubDate>Sun, 11 May 2025 03:12:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=43951043</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=43951043</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43951043</guid></item><item><title><![CDATA[New comment by mkaic in "Cautious Optimizers: Improving Training with One Line of Code"]]></title><description><![CDATA[
<p>Damn, this is a strikingly simple modification. Basically, modern deep learning optimizers typically calculate the update to the weights each step using some kind of momentum and/or LR scaling based on the running variance of the gradients. This means that, in theory, the actual "instantaneous" gradients from a particular backward pass might point in a different direction than the actual update the optimizer applies. The change the authors propose is to simply ignore any parameter updates proposed by the optimizer that have the opposite sign of the current gradient from the most recent backwards pass. They're essentially saying "only apply the long-term stabilized update where it <i>agrees</i> with the current 'instantaneous' gradient." They show that this simple change significantly speeds up model training.<p>I'm pretty intrigued by this, but will, as usual, wait for independent replications to come out before I fully believe it. That said, because of how simple this is, I'd expect such replications to happen within 24 hours. Exciting work!</p>
]]></description><pubDate>Mon, 03 Mar 2025 20:16:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=43246161</link><dc:creator>mkaic</dc:creator><comments>https://news.ycombinator.com/item?id=43246161</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43246161</guid></item></channel></rss>