{"id":280389,"date":"2024-05-07T04:16:18","date_gmt":"2024-05-07T11:16:18","guid":{"rendered":"https:\/\/sftarticles.wpenginepowered.com\/es\/?p=331839"},"modified":"2025-07-01T16:33:45","modified_gmt":"2025-07-01T23:33:45","slug":"this-robot-does-better-yoga-than-you","status":"publish","type":"post","link":"https:\/\/cms-articles.softonic.io\/en\/this-robot-does-better-yoga-than-you\/","title":{"rendered":"This robot does better yoga than you"},"content":{"rendered":"\n<p>This four-legged creature balancing on an exercise ball is a fun experiment to watch, but <strong>ultimately it demonstrates that<\/strong> <strong>AIs like GPT-4 can train robots to perform complex tasks<\/strong> in the real world much more effectively than humans.<\/p>\n\n\n<div class=\"sc-card-program\">\r\n  <div class=\"sc-card-program__body\">\r\n    <div class=\"sc-card-program__row clearfix\">\r\n      <div class=\"sc-card-program__col-logo\">\r\n        <img decoding=\"async\" class=\"sc-card-program__img\" alt=\"ChatGPT\" src=\"https:\/\/images.sftcdn.net\/images\/t_app-icon-s\/p\/47ef1772-2a82-4750-b97a-354b13dbd112\/3647786732\/chatgpt-ChatGPT-icon.png\" width=\"100px\" height=\"100px\">\r\n      <\/div>\r\n      <div class=\"sc-card-program__col-title\">\r\n        <span class=\"sc-card-program__title\">ChatGPT<\/span>\r\n        <a class=\"sc-card-program__button sc-card-program-internal\" href=\"https:\/\/chatgpt.en.softonic.com\/iphone\" target=\"_self\" rel=\"noopener noreferrer\">DOWNLOAD<\/a>\r\n      <\/div>\r\n      <div class=\"sc-card-program__col-rating\">\r\n        <svg class=\"rating-score__content\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" version=\"1.1\" x=\"0\" y=\"0\" viewbox=\"0 0 50 50\" enable-background=\"new 0 0 50 50\" xml:space=\"preserve\"><path class=\"rating-score__background rating-score--good\" fill=\"none\" stroke-width=\"6\" stroke-miterlimit=\"10\" d=\"M40 40c8.3-8.3 8.3-21.7 0-30s-21.7-8.3-30 0 -8.3 21.7 0 30\"><\/path><path class=\"rating-score__value rating-score__value--0\" fill=\"none\" stroke-width=\"6\" stroke-dashoffset=\"0\" stroke-miterlimit=\"10\" d=\"M40 40c8.3-8.3 8.3-21.7 0-30s-21.7-8.3-30 0 -8.3 21.7 0 30\"><\/path><text class=\"rating-score__number\" content=\"\" text-anchor=\"middle\" transform=\"matrix(1 0 0 1 25 31.0837)\" data-auto=\"app-user-score\"><\/text><\/svg>\r\n      <\/div>\r\n    <\/div>\r\n    <div class=\"sc-card-program__row\">\r\n      <span class=\"sc-card-program__description\"><\/span>\r\n    <\/div>\r\n    <div class=\"sc-card-program__row\">\r\n      <img decoding=\"async\" class=\"sc-card-program__bigpic\" src=\"\" onerror=\"this.style.display='none'\">\r\n    <\/div>\r\n    <a class=\"sc-card-program__link track-link sc-card-program-internal\" href=\"https:\/\/chatgpt.en.softonic.com\/iphone\" target=\"_self\" rel=\"noopener noreferrer\"><\/a>\r\n  <\/div>\r\n<\/div>\n\n\n\n<p>DrEureka, a new open-source software package, is used to train robots to perform real-world tasks <a href=\"https:\/\/en.softonic.com\/articles\/openai-has-just-fixed-the-biggest-problem-of-gpt-4\" target=\"_blank\" rel=\"noopener\" title=\"using large language models (LLM) like ChatGPT 4.\">using large language models (LLM) like ChatGPT 4.<\/a> <strong>It is a &#8220;sim-to-reality&#8221; system, which means it teaches robots in a virtual environment using simulated physics before deploying them in the physical space.<\/strong><\/p>\n\n\n\n<p>Dr. Jim Fan, one of the developers of DrEureka, used a quadruped robot Unitree Go1 to make headlines. <strong>It is a &#8220;low-cost&#8221; and open-source robot, which is very useful because even with AI, robot pets are prone to falling<\/strong>. As for the &#8220;low-cost&#8221; part, well&#8230; It is sold on Amazon for $5,899 and has a 1-star rating.<\/p>\n\n\n\n<figure class=\"wp-block-embed is-type-rich is-provider-twitter wp-block-embed-twitter\"><div class=\"wp-block-embed__wrapper\">\n<blockquote class=\"twitter-tweet\" data-width=\"550\" data-dnt=\"true\"><p lang=\"en\" dir=\"ltr\">We trained a robot dog to balance and walk on top of a yoga ball purely in simulation, and then transfer zero-shot to the real world. No fine-tuning. Just works. <br><br>I\u2019m excited to announce DrEureka, an LLM agent that writes code to train robot skills in simulation, and writes more\u2026 <a href=\"https:\/\/t.co\/kuG14LmSOh\">pic.twitter.com\/kuG14LmSOh<\/a><\/p>&mdash; Jim Fan (@DrJimFan) <a href=\"https:\/\/twitter.com\/DrJimFan\/status\/1786429467537088741?ref_src=twsrc%5Etfw\">May 3, 2024<\/a><\/blockquote><script async src=\"https:\/\/platform.twitter.com\/widgets.js\" charset=\"utf-8\"><\/script>\n<\/div><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">Robots that are better than us in almost everything<\/h2>\n\n\n\n<p>The &#8220;Dr&#8221; in DrEureka stands for &#8220;Domain randomization&#8221;, which means randomizing variables such as friction, mass, damping, center of gravity, etc. in a simulated environment.<\/p>\n\n\n\n<p><strong>With a few instructions in an LLM like ChatGPT, AI can write code <\/strong>that creates a reward\/punishment system to train the robot in the virtual space, where 0 = failure, and anything above 0 is a victory. The higher the score, the better.<\/p>\n\n\n\n<p>You can create parameters by minimizing and maximizing failure\/breakage points in areas such as ball bounce, driving force, degree of freedom of your limbs, and damping, to name a few<strong>. As LLM, you have no problem creating them in large volumes for the training system to execute them simultaneously.<\/strong><\/p>\n\n\n\n<p>After each simulation, <strong>GPT can also reflect on how well the virtual robot has performed and how it can improve<\/strong>. Exceeding or violating parameters, for example, overheating a motor or attempting to articulate a limb beyond its capabilities, will result in a 0&#8230; And no one likes to get zero points, not even an AI.<\/p>\n\n\n<div class=\"sc-card-program\">\r\n  <div class=\"sc-card-program__body\">\r\n    <div class=\"sc-card-program__row clearfix\">\r\n      <div class=\"sc-card-program__col-logo\">\r\n        <img decoding=\"async\" class=\"sc-card-program__img\" alt=\"ChatGPT\" src=\"https:\/\/images.sftcdn.net\/images\/t_app-icon-s\/p\/47ef1772-2a82-4750-b97a-354b13dbd112\/3647786732\/chatgpt-ChatGPT-icon.png\" width=\"100px\" height=\"100px\">\r\n      <\/div>\r\n      <div class=\"sc-card-program__col-title\">\r\n        <span class=\"sc-card-program__title\">ChatGPT<\/span>\r\n        <a class=\"sc-card-program__button sc-card-program-internal\" href=\"https:\/\/chatgpt.en.softonic.com\/iphone\" target=\"_self\" rel=\"noopener noreferrer\">DOWNLOAD<\/a>\r\n      <\/div>\r\n      <div class=\"sc-card-program__col-rating\">\r\n        <svg class=\"rating-score__content\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" version=\"1.1\" x=\"0\" y=\"0\" viewbox=\"0 0 50 50\" enable-background=\"new 0 0 50 50\" xml:space=\"preserve\"><path class=\"rating-score__background rating-score--good\" fill=\"none\" stroke-width=\"6\" stroke-miterlimit=\"10\" d=\"M40 40c8.3-8.3 8.3-21.7 0-30s-21.7-8.3-30 0 -8.3 21.7 0 30\"><\/path><path class=\"rating-score__value rating-score__value--0\" fill=\"none\" stroke-width=\"6\" stroke-dashoffset=\"0\" stroke-miterlimit=\"10\" d=\"M40 40c8.3-8.3 8.3-21.7 0-30s-21.7-8.3-30 0 -8.3 21.7 0 30\"><\/path><text class=\"rating-score__number\" content=\"\" text-anchor=\"middle\" transform=\"matrix(1 0 0 1 25 31.0837)\" data-auto=\"app-user-score\"><\/text><\/svg>\r\n      <\/div>\r\n    <\/div>\r\n    <div class=\"sc-card-program__row\">\r\n      <span class=\"sc-card-program__description\"><\/span>\r\n    <\/div>\r\n    <div class=\"sc-card-program__row\">\r\n      <img decoding=\"async\" class=\"sc-card-program__bigpic\" src=\"\" onerror=\"this.style.display='none'\">\r\n    <\/div>\r\n    <a class=\"sc-card-program__link track-link sc-card-program-internal\" href=\"https:\/\/chatgpt.en.softonic.com\/iphone\" target=\"_self\" rel=\"noopener noreferrer\"><\/a>\r\n  <\/div>\r\n<\/div>\n\n\n\n<p><strong>How did he do? Better than us. <\/strong>DrEureka was able to outperform humans in robot training, with a 34% advantage in speed and a 20% advantage in distance covered on real-world mixed terrains.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>After each simulation, GPT can also reflect on how well the virtual robot has performed and how it can improve. This robot is incredible.<\/p>\n","protected":false},"author":9265,"featured_media":280391,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","wpcf-pageviews":1},"categories":[1015],"tags":[1604],"usertag":[],"vertical":[],"content-category":[],"class_list":["post-280389","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-robot"],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts\/280389","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/users\/9265"}],"replies":[{"embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/comments?post=280389"}],"version-history":[{"count":1,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts\/280389\/revisions"}],"predecessor-version":[{"id":313348,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/posts\/280389\/revisions\/313348"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/media\/280391"}],"wp:attachment":[{"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/media?parent=280389"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/categories?post=280389"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/tags?post=280389"},{"taxonomy":"usertag","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/usertag?post=280389"},{"taxonomy":"vertical","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/vertical?post=280389"},{"taxonomy":"content-category","embeddable":true,"href":"https:\/\/cms-articles.softonic.io\/en\/wp-json\/wp\/v2\/content-category?post=280389"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}