{"id":794,"date":"2025-01-29T00:39:42","date_gmt":"2025-01-29T00:39:42","guid":{"rendered":"https:\/\/www.insuracarelife.com\/blog\/do-deepseeks-a-i-advances-mean-us-tech-controls-have-failed\/"},"modified":"2025-01-29T00:39:42","modified_gmt":"2025-01-29T00:39:42","slug":"do-deepseeks-a-i-advances-mean-us-tech-controls-have-failed","status":"publish","type":"post","link":"https:\/\/www.insuracarelife.com\/blog\/do-deepseeks-a-i-advances-mean-us-tech-controls-have-failed\/","title":{"rendered":"Do DeepSeek\u2019s A.I. Advances Mean US Tech Controls Have Failed?"},"content":{"rendered":"<div data-testid=\"companionColumn-0\">\n<div class=\"css-53u6y8\">\n<p class=\"css-at9mc1 evys1bk0\">The United States has worked steadily over the past three years to limit China\u2019s access to the cutting edge computer chips that power advanced artificial intelligence systems. Its aim has been to slow China\u2019s progress in developing sophisticated A.I. models.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Now a Chinese firm, DeepSeek, has created that very technology. In recent weeks, DeepSeek released multiple A.I. models and a chatbot whose performance rivals that of the best products made by American firms, all while using far fewer of the high-cost A.I. chips that companies typically need. Over the weekend, DeepSeek\u2019s chatbot shot to the top of Apple\u2019s App Store charts as people downloaded it around the world.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">The development has raised big questions about export controls built by the United States in recent years. The Biden administration set up a system of global rules and steadily expanded them to try to keep advanced A.I. technology \u2014 particularly chips made by Nvidia \u2014 out of Chinese hands. They were concerned that technology would give China an edge not just economically, but also militarily.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">DeepSeek\u2019s development has provoked a fierce debate over whether U.S. technology controls have failed. Here\u2019s what to know.<\/p>\n<\/div>\n<aside class=\"css-ew4tgv\" aria-label=\"companion column\"\/><\/div>\n<div data-testid=\"companionColumn-1\">\n<div class=\"css-53u6y8\">\n<h3 class=\"css-15h6bi9 e1gnsphs0\" id=\"link-106af6fa\"><span><strong class=\"css-8qgvsz ebyp5n10\">DeepSeek\u2019s innovations suggest the Biden administration may have acted too slowly to keep up with private companies sidestepping its controls.<\/strong><\/span><\/h3>\n<p class=\"css-at9mc1 evys1bk0\">DeepSeek has said that its most recent model <a class=\"css-yywogo\" href=\"https:\/\/arxiv.org\/abs\/2412.19437\" title=\"\" rel=\"noopener noreferrer\" target=\"_blank\">was trained on<\/a> Nvidia H800s. This is an A.I. chip that Nvidia developed specifically for the Chinese market after export controls were first imposed, and that caused a fair amount of drama in Washington.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">When the United States put restrictions on Nvidia\u2019s most advanced chips in 2022, Nvidia quickly adapted by creating slightly downgraded chips that fell just under the threshold the government had set. These chips were technically legal for Chinese companies to use, but allowed them to achieve practically the same results.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">This angered Biden officials, and they moved to restrict the new chips as well. But the government moved slowly, and it took them about a year to ban the H800 and other downgraded chips. In the meantime, Chinese companies stockpiled a lot of them.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">It\u2019s not clear how DeepSeek obtained its Nvidia H800s, but it would have been legal for the company to buy them in late 2022 or 2023. Now, however, such purchases would not be.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">\u201cYou can\u2019t control what\u2019s already there,\u201d said<strong class=\"css-8qgvsz ebyp5n10\"> <\/strong>Jimmy Goodrich, a senior adviser for technology analysis at the RAND Corporation. \u201cHad the Biden administration more quickly responded and limited the H800 to China, there\u2019s no doubt DeepSeek would have been more challenged in putting this model out.\u201d<\/p>\n<\/div>\n<aside class=\"css-ew4tgv\" aria-label=\"companion column\"\/><\/div>\n<div data-testid=\"companionColumn-2\">\n<div class=\"css-53u6y8\">\n<p class=\"css-at9mc1 evys1bk0\">DeepSeek also spent years building up its chip supply before Washington\u2019s controls took effect. By 2021, DeepSeek was one of just a handful of Chinese companies that had acquired at least 10,000 Nvidia A100s, the advanced chip Nvidia released in 2020, according to <a class=\"css-yywogo\" href=\"https:\/\/mp.weixin.qq.com\/s\/Cajwfve7f-z2Blk9lnD0hA\" title=\"\" rel=\"noopener noreferrer\" target=\"_blank\">an interview<\/a> with Liang Wenfeng, the founder of DeepSeek, in the Chinese media outlet 36Kr.<\/p>\n<h3 class=\"css-15h6bi9 e1gnsphs0\" id=\"link-29d6f23e\"><span><strong class=\"css-8qgvsz ebyp5n10\">The U.S. has also struggled to stamp out chip smuggling.<\/strong><\/span><\/h3>\n<p class=\"css-at9mc1 evys1bk0\">There\u2019s no evidence that DeepSeek has used smuggled chips. But many Chinese A.I. companies have. Alexandr Wang, the chief executive of the A.I. training giant Scale AI, told The New York Times that Chinese companies had far more high-end chips than U.S. restrictions allowed, and that DeepSeek probably had about 50,000 Nvidia advanced H100 processors, \u201cwhich they obviously can\u2019t talk about.\u201d<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Both Nvidia and the U.S. government have argued that the scale of smuggling was limited. But The Times last year reported an active trade in China in restricted A.I. technology. In a bustling market in Shenzhen, in southern China, chip vendors reported engaging in sales involving hundreds or thousands of restricted chips.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Representatives of 11 companies said they sold or transported banned Nvidia chips \u2014 including A100s and H100s, the company\u2019s most advanced at the time \u2014 and The Times found dozens more businesses offering them online. One vendor in Shenzhen showed a reporter screenshots arranging deliveries of servers containing more than 2,000 of Nvidia\u2019s most advanced chips, a transaction totaling $103 million.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Since then, <a class=\"css-yywogo\" href=\"https:\/\/www.theinformation.com\/articles\/nvidia-ai-chip-smuggling-to-china-becomes-an-industry\" title=\"\" rel=\"noopener noreferrer\" target=\"_blank\">more reports<\/a> have emerged documenting large-scale smuggling, particularly through other countries in Asia.<\/p>\n<\/div>\n<aside class=\"css-ew4tgv\" aria-label=\"companion column\"\/><\/div>\n<div data-testid=\"companionColumn-3\">\n<div class=\"css-53u6y8\">\n<p class=\"css-at9mc1 evys1bk0\">The Biden administration released a sweeping regulation this month that aims to deal with the smuggling issue, by setting caps on the number of chips that Nvidia can sell to every country worldwide.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">It remains to be seen what the Trump administration will do about it. In a trade executive order President Trump signed on his first day in office, however, he ordered his officials to review the U.S. export control system, including \u201chow to identify and eliminate loopholes in existing export controls.\u201d<\/p>\n<h3 class=\"css-15h6bi9 e1gnsphs0\" id=\"link-19a2ce96\"><span>U.S. controls appear to have encouraged Chinese ingenuity \u2014 but they have also clearly held back China\u2019s A.I. development.<\/span><\/h3>\n<p class=\"css-at9mc1 evys1bk0\">American technology restrictions appear to have accelerated the efforts of Chinese researchers to try to do more with less.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">The most notable thing about DeepSeek\u2019s model is that, according to the company, it was developed with just a fraction of the high-priced chips that Western companies have used to make similar technology. DeepSeek\u2019s engineers said they used only about 2,000 Nvidia chips, whereas most top companies have trained chatbots using 16,000 chips or more. Nvidia\u2019s shares plunged sharply on Monday on fears that technology companies will be able to do cutting-edge A.I. in the future while paying Nvidia far less.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Jeffrey Ding, a professor at George Washington University who studies emerging technologies, said that most global companies have been using ever-larger amounts of computing power and data to improve A.I. performance. But DeepSeek and other Chinese firms had been \u201cforced to go down this other pathway to find out whether we can get good enough performance with lower training costs and less compute,\u201d he said.<\/p>\n<\/div>\n<aside class=\"css-ew4tgv\" aria-label=\"companion column\"\/><\/div>\n<div data-testid=\"companionColumn-4\">\n<div class=\"css-53u6y8\">\n<p class=\"css-at9mc1 evys1bk0\">The implications of cheaper models like DeepSeek\u2019s could be profound. With DeepSeek openly sharing details about how it built its model, companies in China and around the world will be able to replicate its low-cost approach.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">That means \u201cit will be much cheaper and could be far less energy intensive for anyone to build and run A.I., from U.S. hyperscalers to Midwestern small businesses, North Korean hackers and Russia\u2019s military,\u201d said Martin Chorzempa, a senior fellow at the Peterson Institute for International Economics.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Still, China would likely be much further ahead in A.I. without the export controls. In interviews, DeepSeek\u2019s founder has acknowledged that the lack of access to computing power was a limitation for the company.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Unlike American A.I. companies, DeepSeek will not be able to legally purchase the newest generation of A.I. chips that Nvidia is rolling out right now, which multiplies the speed and performance of the previous chips.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">\u201cAnyone worried about what DeepSeek can do today would be more worried if they had done it with access to the far superior computing resources their U.S. competitors have,\u201d Mr. Chorzempa said.<\/p>\n<\/div>\n<aside class=\"css-ew4tgv\" aria-label=\"companion column\"\/><\/div>\n<div data-testid=\"companionColumn-5\">\n<div class=\"css-53u6y8\">\n<p class=\"css-at9mc1 evys1bk0\">DeepSeek\u2019s success suggests that Silicon Valley\u2019s lead on A.I. has shrunk, despite efforts by Washington to limit Chinese access to the advanced chips. But it\u2019s notable that DeepSeek is still building its models on Nvidia chips \u2014 not on the rival A.I. chips that the Chinese technology firm Huawei is trying to develop.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">Some Chinese computer engineers have suggested it would be possible to run the latest DeepSeek model on a larger number of less advanced chips, including those made by Huawei, even though Huawei\u2019s A.I. chips are much lower performing.<\/p>\n<p class=\"css-at9mc1 evys1bk0\">But no Chinese company is yet able to make advanced A.I. chips that rival Nvidia\u2019s, or makes the type of complex machinery needed to make those chips. \u201cThe only advantage the United States still has over China at this moment is in hardware,\u201d Mr. Goodrich said.<\/p>\n<\/div>\n<aside class=\"css-ew4tgv\" aria-label=\"companion column\"\/><\/div>\n","protected":false},"excerpt":{"rendered":"<p>The United States has worked steadily over the past three years to limit China\u2019s access to the cutting edge computer chips that power advanced artificial intelligence systems. Its aim has been to slow China\u2019s progress in developing sophisticated A.I. models. Now a Chinese firm, DeepSeek, has created that very technology. In recent weeks, DeepSeek released [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":795,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[514,515,517,513,518,516],"class_list":["post-794","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog","tag-a-i","tag-advances","tag-controls","tag-deepseeks","tag-failed","tag-tech"],"_links":{"self":[{"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/posts\/794","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/comments?post=794"}],"version-history":[{"count":0,"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/posts\/794\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/media\/795"}],"wp:attachment":[{"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/media?parent=794"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/categories?post=794"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.insuracarelife.com\/blog\/wp-json\/wp\/v2\/tags?post=794"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}