<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: siekmanj</title><link>https://news.ycombinator.com/user?id=siekmanj</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 30 Jul 2026 23:27:36 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=siekmanj" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by siekmanj in "Gemini Robotics 2 brings whole body intelligence to robots"]]></title><description><![CDATA[
<p>Actuators on modern humanoids use cycloids or planetary reducers, not strain waves. They are much more compliant and backdrivable because of the huge reduction in reflected inertia.</p>
]]></description><pubDate>Thu, 30 Jul 2026 19:24:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=49114508</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=49114508</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49114508</guid></item><item><title><![CDATA[New comment by siekmanj in "Gemini Robotics 2 brings whole body intelligence to robots"]]></title><description><![CDATA[
<p>I could not disagree more with the claim about no improvement in robotic actuators.<p>The torque density and price of actuators has fallen dramatically since Ben Katz's MIT work on mini cheetah. The actuators on the Unitree G1 based on that work are powerful for their size and near quasi-direct-drive. The motors on the BD E-Atlas are completely passively cooled and appear to have really good torque density. Actuators have never been improving faster than they are now.</p>
]]></description><pubDate>Thu, 30 Jul 2026 18:55:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=49114143</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=49114143</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49114143</guid></item><item><title><![CDATA[Agility’s Latest Digit Robot Prepares for Its First Job]]></title><description><![CDATA[
<p>Article URL: <a href="https://spectrum.ieee.org/agility-robotics-digit">https://spectrum.ieee.org/agility-robotics-digit</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=35233997">https://news.ycombinator.com/item?id=35233997</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 20 Mar 2023 16:25:32 +0000</pubDate><link>https://spectrum.ieee.org/agility-robotics-digit</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=35233997</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=35233997</guid></item><item><title><![CDATA[New comment by siekmanj in "Ask HN: What were the papers on the list Ilya Sutskever gave John Carmack?"]]></title><description><![CDATA[
<p>"RL: A Deep Reinforcement Learning Framework" seems to have been hallucinated, does not exist.</p>
]]></description><pubDate>Fri, 03 Feb 2023 17:59:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=34645140</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=34645140</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34645140</guid></item><item><title><![CDATA[New comment by siekmanj in "Bipedal robot developed at Oregon State achieves 100m Guinness World Record"]]></title><description><![CDATA[
<p>Video here: <a href="https://twitter.com/oregonstatenews/status/1574796924326010880" rel="nofollow">https://twitter.com/oregonstatenews/status/15747969243260108...</a></p>
]]></description><pubDate>Tue, 27 Sep 2022 20:01:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=33000377</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=33000377</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33000377</guid></item><item><title><![CDATA[Bipedal robot developed at Oregon State achieves 100m Guinness World Record]]></title><description><![CDATA[
<p>Article URL: <a href="https://today.oregonstate.edu/news/bipedal-robot-developed-oregon-state-achieves-guinness-world-record-100-meters">https://today.oregonstate.edu/news/bipedal-robot-developed-oregon-state-achieves-guinness-world-record-100-meters</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=33000357">https://news.ycombinator.com/item?id=33000357</a></p>
<p>Points: 7</p>
<p># Comments: 1</p>
]]></description><pubDate>Tue, 27 Sep 2022 19:59:20 +0000</pubDate><link>https://today.oregonstate.edu/news/bipedal-robot-developed-oregon-state-achieves-guinness-world-record-100-meters</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=33000357</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33000357</guid></item><item><title><![CDATA[New comment by siekmanj in "The biggest threat facing middle-age men is loneliness (2017)"]]></title><description><![CDATA[
<p><a href="https://archive.ph/8X76D" rel="nofollow">https://archive.ph/8X76D</a></p>
]]></description><pubDate>Mon, 18 Jul 2022 15:22:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=32139416</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=32139416</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=32139416</guid></item><item><title><![CDATA[New comment by siekmanj in "Python 3.11 vs 3.10 performance"]]></title><description><![CDATA[
<p>True, but it's not trying to be multi-threaded, just concurrent.</p>
]]></description><pubDate>Wed, 06 Jul 2022 15:43:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=32002617</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=32002617</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=32002617</guid></item><item><title><![CDATA[New comment by siekmanj in "Python 3.11 vs 3.10 performance"]]></title><description><![CDATA[
<p>Python has actually had concurrency since about 2019: <a href="https://docs.python.org/3/library/asyncio.html" rel="nofollow">https://docs.python.org/3/library/asyncio.html</a>. Having used it a few times, it seems fairly sane, but tbf my experience with concurrency in other languages is fairly limited.<p>edit: ray <a href="https://github.com/ray-project/ray" rel="nofollow">https://github.com/ray-project/ray</a> is also pretty easy to use and powerful for actual parallelism</p>
]]></description><pubDate>Wed, 06 Jul 2022 15:24:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=32002346</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=32002346</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=32002346</guid></item><item><title><![CDATA[Agility Robotics Raises $150M]]></title><description><![CDATA[
<p>Article URL: <a href="https://agilityrobotics.com/news/2022/future-robotics">https://agilityrobotics.com/news/2022/future-robotics</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=31122560">https://news.ycombinator.com/item?id=31122560</a></p>
<p>Points: 1</p>
<p># Comments: 0</p>
]]></description><pubDate>Fri, 22 Apr 2022 15:10:40 +0000</pubDate><link>https://agilityrobotics.com/news/2022/future-robotics</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=31122560</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=31122560</guid></item><item><title><![CDATA[New comment by siekmanj in "Finland or Sweden joining NATO would spark Russian response, Russia warns"]]></title><description><![CDATA[
<p>From some cursory googling:<p>Afghanistan peak US troops: 102,000 (2011)<p>Gulf war peak US troops: 700,000 (1990-1991)<p>Iraq peak US troops: range from 168,000 - 192,000 (2007)<p>Vietnam peak US troops: 543,000 (1969)<p>Korea peak US troops: 320,000, unclear what year<p>edit: obviously, Russia's involvement in Ukraine over the last few days would be by far the biggest operation in Europe since WW2. Just not globally (by a long shot).</p>
]]></description><pubDate>Fri, 25 Feb 2022 19:44:33 +0000</pubDate><link>https://news.ycombinator.com/item?id=30471427</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=30471427</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=30471427</guid></item><item><title><![CDATA[New comment by siekmanj in "Open-sourcing the MuJoCo physics simulator for robotics"]]></title><description><![CDATA[
<p>This is an amazing outcome. The license cost was a huge barrier to entry and the progress of deep RL generally.</p>
]]></description><pubDate>Mon, 18 Oct 2021 19:30:36 +0000</pubDate><link>https://news.ycombinator.com/item?id=28910395</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=28910395</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=28910395</guid></item><item><title><![CDATA[New comment by siekmanj in "Autonomous drone navigates independently through the forest at 40 km/h"]]></title><description><![CDATA[
<p>The sim2real approach can work pretty well as long as you are <i>very</i> in tune with where your simulator falls short relative to the real world and take steps to circumvent those shortcomings.<p>Here's some work my lab did on sim2real for a roughly human-scale bipedal robot (Cassie): <a href="https://www.youtube.com/watch?v=MPhEmC6b6XU" rel="nofollow">https://www.youtube.com/watch?v=MPhEmC6b6XU</a><p>We were able to train the robot to climb stairs completely by feel/proprioception without any sort of vision. We trained it in simulation, and then transferred it to the real world without issue.</p>
]]></description><pubDate>Fri, 08 Oct 2021 03:57:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=28795592</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=28795592</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=28795592</guid></item><item><title><![CDATA[New comment by siekmanj in "RMA: Rapid Motor Adaptation for Legged Robots"]]></title><description><![CDATA[
<p>Since you mentioned it--here's what Oregon State University and the Cassie group is up to these days: <a href="https://youtu.be/MPhEmC6b6XU" rel="nofollow">https://youtu.be/MPhEmC6b6XU</a></p>
]]></description><pubDate>Tue, 13 Jul 2021 05:22:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=27817826</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=27817826</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27817826</guid></item><item><title><![CDATA[New comment by siekmanj in "EleutherAI One Year Retrospective"]]></title><description><![CDATA[
<p>It's a German phrase which colloquially translated means 'Gotta go bro'</p>
]]></description><pubDate>Thu, 08 Jul 2021 21:52:30 +0000</pubDate><link>https://news.ycombinator.com/item?id=27778061</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=27778061</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27778061</guid></item><item><title><![CDATA[New comment by siekmanj in "Thinking Like Transformers"]]></title><description><![CDATA[
<p>Actually, I think it is conventional neural networks which can only approximate finite state machines. RNNs are (in theory, not so much in practice) Turing complete.</p>
]]></description><pubDate>Wed, 16 Jun 2021 18:55:21 +0000</pubDate><link>https://news.ycombinator.com/item?id=27532111</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=27532111</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27532111</guid></item><item><title><![CDATA[New comment by siekmanj in "Agility Robotics' Cassie Is Now Astonishingly Good at Stairs"]]></title><description><![CDATA[
<p>Hey everyone, I am one of the authors of the paper described in this article. We used reinforcement learning and recurrent neural networks to learn a controller which can ascend and descend stairs without any vision-based perception, meaning that it must rely entirely on proprioception to walk. This is the first time (to our knowledge) that a human-sized bipedal robot has been able to climb real-world stairs blind to the world, and we're pretty excited about the results of this research.<p>Here's a link to the Arxiv submission: <a href="https://arxiv.org/abs/2105.08328" rel="nofollow">https://arxiv.org/abs/2105.08328</a><p>And here's the accompanying video: 
<a href="https://youtu.be/MPhEmC6b6XU" rel="nofollow">https://youtu.be/MPhEmC6b6XU</a><p>And an uninterrupted five-minute video of a test on an outdoor staircase:
<a href="https://youtu.be/nuhHiKEtaZQ" rel="nofollow">https://youtu.be/nuhHiKEtaZQ</a><p>Happy to answer any questions the HN crowd may have!</p>
]]></description><pubDate>Thu, 20 May 2021 22:11:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=27228062</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=27228062</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27228062</guid></item><item><title><![CDATA[Agility Robotics' Cassie Is Now Astonishingly Good at Stairs]]></title><description><![CDATA[
<p>Article URL: <a href="https://spectrum.ieee.org/automaton/robotics/humanoids/agility-robotics-cassie-stairs">https://spectrum.ieee.org/automaton/robotics/humanoids/agility-robotics-cassie-stairs</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=27227730">https://news.ycombinator.com/item?id=27227730</a></p>
<p>Points: 3</p>
<p># Comments: 1</p>
]]></description><pubDate>Thu, 20 May 2021 21:43:25 +0000</pubDate><link>https://spectrum.ieee.org/automaton/robotics/humanoids/agility-robotics-cassie-stairs</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=27227730</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27227730</guid></item><item><title><![CDATA[Watch a 'blind' robot successfully navigate stairs]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.engadget.com/a-bipedal-robot-has-learned-to-tackle-stairs-blind-102647481.html">https://www.engadget.com/a-bipedal-robot-has-learned-to-tackle-stairs-blind-102647481.html</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=27210438">https://news.ycombinator.com/item?id=27210438</a></p>
<p>Points: 2</p>
<p># Comments: 1</p>
]]></description><pubDate>Wed, 19 May 2021 15:59:14 +0000</pubDate><link>https://www.engadget.com/a-bipedal-robot-has-learned-to-tackle-stairs-blind-102647481.html</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=27210438</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=27210438</guid></item><item><title><![CDATA[New comment by siekmanj in "Facebook has 10k engineers working on AR/VR"]]></title><description><![CDATA[
<p>I was always very unimpressed with VR until I borrowed a headset to play Half Life Alyx. That game convinced me VR is the inevitable future of gaming. The rest of the industry may not be there yet, but it is an obviously superior experience.</p>
]]></description><pubDate>Mon, 15 Mar 2021 07:11:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=26463291</link><dc:creator>siekmanj</dc:creator><comments>https://news.ycombinator.com/item?id=26463291</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=26463291</guid></item></channel></rss>