CoinInsight360.com logo CoinInsight360.com logo
A company that is changing the way the world mines bitcoin

WallStreet Forex Robot 3.0
Cryptopolitan 2025-01-28 13:22:37

DeepSeek is unlike anything OpenAI has seen despite Sam Altman’s confidence

DeepSeek AI V3, a Chinese-based application, is now the top-rated free artificial intelligence (AI) app in the United States of America! And as much as Sam Altman and his OpenAI family are playing it “calm” and calling the new model “impressive,” they are getting battered in a game they dominated until last weekend. The US vs. China or DeepSeek vs OpenAI sweepstakes is a heated race to occupy the head of the AI table. But is Sam Altman really doing enough for the American market? Well, if users, especially from the US , feel that an open-source tech model using a “favorable” computational power and has been around for 2 weeks is better than ChatGPT’s -4.0, Altman is in trouble. On January 23, ChatGPT went down for a couple of hours; jokes flew around social media that “productivity went down” too. At that moment in time, ‘competition’ was just a word in the English dictionary for OpenAI. A lot has changed since then. A Chinese company’s AI product is at the top of the mountain in the US. DeepSeek R1: Fear of the open-source unknown In late 2022, OpenAI debuted ChatGPT, much to the dismay of a Chinese AI market that was not ready. Developers in the Asian country rushed to make half-baked chatbots, trying to match the tech giant’s capable technology. None of the products they released, including Baidu, was enough to satisfy users’ requirements and needs. A little over three months after December 2022, DeepSeek was founded by AI hedge fund High Flyer co-founder Liang Wenfeng. Initially rooted in High-Flyer’s AI research unit, the startup transitioned in April 2023 to concentrate on large language models and the pursuit of artificial general intelligence (AGI)—a level of AI capability that matches or exceeds human intelligence across diverse tasks. Very little is known about Wenfeng; users typically don’t really need to know much about the “who.” They care more about results, and in a side-by-side comparison, DeepSeek’s models are blowing ChatGPT out of the water. Unlike OpenAI’s ChatGPT, DeepSeek R1 fulfills a promise that Sam Altman’s OpenAI once championed; open AI. Not even the storm of Former President Biden’s approved restrictions on chip exports to China in the last three years stopped DeepSeek from training its latest models. The US wants to be the epicenter of everything, and they’ve tried so hard to thwart anything that would help China take the lead. Per a recent CNBC interview , Scale AI CEO Alexandr Wang alleged that DeepSeek possesses 50,000 Nvidia H100 chips, which he did not provide any evidence to support. “What we’ve found is that DeepSeek … is the top performing, or roughly on par with the best American models,” the CEO commented. Wang also said that disclosing the details would violate Washington’s call to ban any exports of such advanced materials to China. While exact figures regarding the cost of building R1 or the number of GPUs involved are not known, analysts at Jefferies suggest they had a training cost of just $5.6 million, far less than what OpenAI, Meta, and most US-based companies budget for similar operations. The mere mention of Nvidia in the whole DeepSeek AI frenzy put the chip-makers in trouble; a 17% stock price wipeout on Monday was all we needed to know how the American markets felt. Does the US government fear China that much? Well, as President Donald Trump told the BBC yesterday, it is a “wake up call.” OpenAI needs to do better In a January 28 post on X, OpenAI CEO Altman said DeepSeek’s R1 is an “impressive model,” for delivering big results at low costs to users. He went on to promise his followers that OpenAI will launch better models and that the company welcomes the competition. “We will obviously deliver much better models and also it’s legit invigorating to have a new competitor! we will pull up some releases,” Altman posted. deepseek's r1 is an impressive model, particularly around what they're able to deliver for the price. we will obviously deliver much better models and also it's legit invigorating to have a new competitor! we will pull up some releases. — Sam Altman (@sama) January 28, 2025 OpenAI charges $7.50 per million input tokens for its o1 model, while DeepSeek offers its R1 model at a 95% discount—just $0.14 per million tokens. In software engineering tasks, DeepSeek R1 narrowly outshines OpenAI, scoring 49.2% on the SWE-bench Verified Resolved test versus OpenAI’s 48.9%. It is also up to 50 times more efficient than leading US AI models like Meta’s Llama 3.1 and Anthropic’s Claude 3.5, using far less computing power to deliver comparable or superior results. Would you blame anyone for forgetting about ChatGPT? Altman’s confident tone is not enough to hide vulnerabilities in OpenAI’s modus operandi. Unlike DeepSeek, OpenAI has resisted fully embracing open-source practices, a stance that has drawn criticism in an industry where nothing is more imperative than transparency and collaboration. Right before our eyes, the assumption that American companies will always lead the AI race is being upended by a more nimble, cost-efficient competitor operating outside the Silicon Valley bubble. If OpenAI, with GPT 5.0 or whatever the next model is named, fails to counter DeepSeek’s open-source tech with something the market will welcome with open arms, then the latter’s dominance is no longer a prediction, it’s a spoiler.

阅读免责声明 : 此处提供的所有内容我们的网站,超链接网站,相关应用程序,论坛,博客,社交媒体帐户和其他平台(“网站”)仅供您提供一般信息,从第三方采购。 我们不对与我们的内容有任何形式的保证,包括但不限于准确性和更新性。 我们提供的内容中没有任何内容构成财务建议,法律建议或任何其他形式的建议,以满足您对任何目的的特定依赖。 任何使用或依赖我们的内容完全由您自行承担风险和自由裁量权。 在依赖它们之前,您应该进行自己的研究,审查,分析和验证我们的内容。 交易是一项高风险的活动,可能导致重大损失,因此请在做出任何决定之前咨询您的财务顾问。 我们网站上的任何内容均不构成招揽或要约