<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: willwhitney</title><link>https://news.ycombinator.com/user?id=willwhitney</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 11 Sep 2026 09:29:09 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=willwhitney" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[Computing Inside an AI]]></title><description><![CDATA[
<p>Article URL: <a href="https://willwhitney.com/computing-inside-ai.html">https://willwhitney.com/computing-inside-ai.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=42413823">https://news.ycombinator.com/item?id=42413823</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 14 Dec 2024 01:18:58 +0000</pubDate><link>https://willwhitney.com/computing-inside-ai.html</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=42413823</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42413823</guid></item><item><title><![CDATA[Learning 3D physics simulators from video]]></title><description><![CDATA[
<p>Article URL: <a href="https://willwhitney.com/learned-simulators.html">https://willwhitney.com/learned-simulators.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=39204418">https://news.ycombinator.com/item?id=39204418</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 31 Jan 2024 14:48:34 +0000</pubDate><link>https://willwhitney.com/learned-simulators.html</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=39204418</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39204418</guid></item><item><title><![CDATA[New comment by willwhitney in "Ask HN: Has Google search become quantitatively worse?"]]></title><description><![CDATA[
<p>Would it be bad if we made advertising... illegal? In small part because it would reduce the principal-agent disparity between what you want to do and what Google wants you to do, but more importantly because it would remove the incentive to create blogspam and game the results.<p>Advertising serves a positive purpose by informing people about products that they will like and otherwise would not find. It also serves a negative purpose by attempting to cause people to make decisions which are not in their best interests.<p>I am genuinely curious what proportion of ads seen serve each of those two purposes.</p>
]]></description><pubDate>Tue, 30 Nov 2021 18:20:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=29395483</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=29395483</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=29395483</guid></item><item><title><![CDATA[Parallelizing neural networks on one GPU with JAX]]></title><description><![CDATA[
<p>Article URL: <a href="http://willwhitney.com/parallel-training-jax.html">http://willwhitney.com/parallel-training-jax.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=25920807">https://news.ycombinator.com/item?id=25920807</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Tue, 26 Jan 2021 20:43:10 +0000</pubDate><link>http://willwhitney.com/parallel-training-jax.html</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=25920807</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=25920807</guid></item><item><title><![CDATA[The New York Times]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.robrhinehart.com/the-new-york-times/">https://www.robrhinehart.com/the-new-york-times/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=25867717">https://news.ycombinator.com/item?id=25867717</a></p>
<p>Points: 3</p>
<p># Comments: 1</p>
]]></description><pubDate>Fri, 22 Jan 2021 02:48:38 +0000</pubDate><link>https://www.robrhinehart.com/the-new-york-times/</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=25867717</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=25867717</guid></item><item><title><![CDATA[New comment by willwhitney in "America's True Unemployment Rate"]]></title><description><![CDATA[
<p>That dip is more than 4M people, so I think it's still unexplained. Perhaps due to a shift in response rates or another artifact of COVID churn in the survey methodology?</p>
]]></description><pubDate>Tue, 20 Oct 2020 15:13:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=24838261</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=24838261</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=24838261</guid></item><item><title><![CDATA[There’s no better time for cities to take space away from cars]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.theverge.com/2020/3/23/21191325/cities-car-free-coronavirus-protected-bike-lanes-air-quality-social-distancing">https://www.theverge.com/2020/3/23/21191325/cities-car-free-coronavirus-protected-bike-lanes-air-quality-social-distancing</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=22668176">https://news.ycombinator.com/item?id=22668176</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 23 Mar 2020 20:06:42 +0000</pubDate><link>https://www.theverge.com/2020/3/23/21191325/cities-car-free-coronavirus-protected-bike-lanes-air-quality-social-distancing</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=22668176</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=22668176</guid></item><item><title><![CDATA[New comment by willwhitney in "Ask HN: Favorite note-taking software?"]]></title><description><![CDATA[
<p>You can actually drag and drop multiple notes into a tag, no find and replace necessary. Minor point though.</p>
]]></description><pubDate>Sun, 15 Jul 2018 13:53:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=17535227</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=17535227</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=17535227</guid></item><item><title><![CDATA[New comment by willwhitney in "Show HN: Unlimited UI design"]]></title><description><![CDATA[
<p>They have a grammatical error ("deliciously looking") in their (2) step.</p>
]]></description><pubDate>Sun, 13 Aug 2017 14:28:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=15003029</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=15003029</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=15003029</guid></item><item><title><![CDATA[New comment by willwhitney in "Beating the World’s Best at Super Smash Bros. with Deep Reinforcement Learning"]]></title><description><![CDATA[
<p>Handling delays (and the uncertainty they entail) is a huge challenge, and I think it'll be a rich area of research. The simplest part of the problem is that delays in action or perception also slow the propagation of reward signals, and credit assignment is still a really hard problem.<p>Thinking further afield, future models could learn to adapt their expectations to fit the behavior of a particular opponent. This kind of metalearning is pretty much a wide open problem, though a pair of (roughly equivalent) papers in this direction recently came out from DeepMind: <a href="https://arxiv.org/abs/1611.05763" rel="nofollow">https://arxiv.org/abs/1611.05763</a> and OpenAI: <a href="https://arxiv.org/abs/1611.02779" rel="nofollow">https://arxiv.org/abs/1611.02779</a> It's going to be really exciting to see how these techniques scale.</p>
]]></description><pubDate>Thu, 23 Feb 2017 00:42:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=13710666</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=13710666</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=13710666</guid></item><item><title><![CDATA[New comment by willwhitney in "Beating the World’s Best at Super Smash Bros. with Deep Reinforcement Learning"]]></title><description><![CDATA[
<p>Yup, it's definitely an advantage to get all the correct values from the game state. But not as much as you might think; the vision portion of a DQN or similar trains quite quickly.<p>Plus, our bot doesn't have any clue about projectiles. We don't know where they live in memory, so the network doesn't get to know about them at all.</p>
]]></description><pubDate>Thu, 23 Feb 2017 00:30:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=13710597</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=13710597</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=13710597</guid></item><item><title><![CDATA[Beating the World’s Best at Super Smash Bros. with Deep Reinforcement Learning]]></title><description><![CDATA[
<p>Article URL: <a href="https://arxiv.org/abs/1702.06230">https://arxiv.org/abs/1702.06230</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=13710253">https://news.ycombinator.com/item?id=13710253</a></p>
<p>Points: 202</p>
<p># Comments: 55</p>
]]></description><pubDate>Wed, 22 Feb 2017 23:27:17 +0000</pubDate><link>https://arxiv.org/abs/1702.06230</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=13710253</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=13710253</guid></item><item><title><![CDATA[New comment by willwhitney in "Sunrise is shutting down"]]></title><description><![CDATA[
<p>Once a startup gets bought, I look for alternatives.<p>If they don't shut down, I still learned something new! But usually they do.<p>In this case, I landed on Fantastical [1] — it's a little pricy but completely worth it. (Mac/iOS only.)<p>[1]: <a href="https://flexibits.com/fantastical" rel="nofollow">https://flexibits.com/fantastical</a></p>
]]></description><pubDate>Wed, 11 May 2016 17:57:55 +0000</pubDate><link>https://news.ycombinator.com/item?id=11677613</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=11677613</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=11677613</guid></item><item><title><![CDATA[Understanding Visual Concepts with Continuation Learning]]></title><description><![CDATA[
<p>Article URL: <a href="http://willwhitney.github.io/understanding-visual-concepts/">http://willwhitney.github.io/understanding-visual-concepts/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=11176851">https://news.ycombinator.com/item?id=11176851</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 25 Feb 2016 19:03:13 +0000</pubDate><link>http://willwhitney.github.io/understanding-visual-concepts/</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=11176851</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=11176851</guid></item><item><title><![CDATA[New comment by willwhitney in "Himawari.js: Download real-time images of Earth from the Himawari-8 satellite"]]></title><description><![CDATA[
<p>I independently wrote a version of this for OS X: <a href="https://github.com/willwhitney/live-earth-desktop" rel="nofollow">https://github.com/willwhitney/live-earth-desktop</a></p>
]]></description><pubDate>Fri, 12 Feb 2016 15:11:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=11087682</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=11087682</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=11087682</guid></item><item><title><![CDATA[Ask HN: Blacklist Forbes?]]></title><description><![CDATA[
<p>Would it be possible to blacklist forbes.com links? They now block viewers who are using ad blockers, and have recently been serving advertising malware to visitors [1].<p>They also have awful interstitial ads [2] and, while they masquerade as a news organization, are actually a blog farm [3].<p>Blocking Forbes would be no great loss in terms of content, and would avoid a surprising amount of advertising- (and malware-) generated pain.<p>[1]: http://www.extremetech.com/internet/220696-forbes-forces-readers-to-turn-off-ad-blockers-promptly-serves-malware<p>[2]: <any Forbes article, e.g. http://www.forbes.com/sites/kathleenchaykowski/2016/01/08/meet-the-queen-of-imgur-the-image-sharing-site-thats-half-the-size-of-twitter/ ><p>[3]: https://np.reddit.com/r/technology/comments/3zs6qk/gq_and_forbes_go_after_ad_blocker_users_rather/cyp2uls?context=3#cyowy51</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=10871410">https://news.ycombinator.com/item?id=10871410</a></p>
<p>Points: 164</p>
<p># Comments: 41</p>
]]></description><pubDate>Sat, 09 Jan 2016 14:53:01 +0000</pubDate><link>https://news.ycombinator.com/item?id=10871410</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=10871410</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=10871410</guid></item><item><title><![CDATA[Popcorn time revived]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.reddit.com/r/PopCornTime/comments/3sbsz6/download_working_installers_or_fix_your_current/">https://www.reddit.com/r/PopCornTime/comments/3sbsz6/download_working_installers_or_fix_your_current/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=10653632">https://news.ycombinator.com/item?id=10653632</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Tue, 01 Dec 2015 04:27:34 +0000</pubDate><link>https://www.reddit.com/r/PopCornTime/comments/3sbsz6/download_working_installers_or_fix_your_current/</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=10653632</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=10653632</guid></item><item><title><![CDATA[New comment by willwhitney in "Ask HN: What do you do if someone is already building what you wanted to build?"]]></title><description><![CDATA[
<p>Help.</p>
]]></description><pubDate>Fri, 19 Jun 2015 18:50:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=9746841</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=9746841</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=9746841</guid></item><item><title><![CDATA[New comment by willwhitney in "Show HN: Hydrogen Brings Light Table to Atom"]]></title><description><![CDATA[
<p>I'd love to do live variable inspection (like the Chrome Dev Tools) — it's something I'm thinking about for the future.<p>There's a (nascent) project called `jmp` (for Jupyter Messaging Protocol) that implements kernel communication in JS: <a href="https://github.com/n-riesco/jmp" rel="nofollow">https://github.com/n-riesco/jmp</a>. I'll probably contribute to that and factor out my kernel communication.</p>
]]></description><pubDate>Thu, 21 May 2015 13:16:46 +0000</pubDate><link>https://news.ycombinator.com/item?id=9582473</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=9582473</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=9582473</guid></item><item><title><![CDATA[New comment by willwhitney in "Show HN: Hydrogen Brings Light Table to Atom"]]></title><description><![CDATA[
<p>:D<p>I mean, if you're offering...</p>
]]></description><pubDate>Wed, 20 May 2015 23:44:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=9580126</link><dc:creator>willwhitney</dc:creator><comments>https://news.ycombinator.com/item?id=9580126</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=9580126</guid></item></channel></rss>