{"id":825826,"date":"2025-02-10T07:12:11","date_gmt":"2025-02-10T13:12:11","guid":{"rendered":"https:\/\/newsycanuse.com\/index.php\/2025\/02\/10\/this-week-in-ai-openai-deep-research-launched-deepseeks-fall\/"},"modified":"2025-02-10T07:12:11","modified_gmt":"2025-02-10T13:12:11","slug":"this-week-in-ai-openai-deep-research-launched-deepseeks-fall","status":"publish","type":"post","link":"https:\/\/newsycanuse.com\/index.php\/2025\/02\/10\/this-week-in-ai-openai-deep-research-launched-deepseeks-fall\/","title":{"rendered":"This Week in AI: OpenAI Deep Research launched; Deepseek\u2019s fall"},"content":{"rendered":"
\n
\n
    \n
  1. Homepage<\/a><\/li>\n
  2. > <\/li>\n
  3. News<\/a><\/li>\n
  4. > <\/li>\n

  5. \n Business
    \n <\/a><\/li>\n
  6. > <\/li>\n
  7. This Week in AI: OpenAI Deep Research launched; Deepseek\u2019s fall<\/li>\n<\/ol>\n<\/div>\n

    OpenAI launches Deep Research: An AI agent for advanced research<\/strong><\/p>\n

    This week, OpenAI<\/a> launched a new artificiaI intelligence (AI) agent<\/a> called Deep Research. Like the first AI agent it released, called Operator, Deep Research<\/a> is an AI agent capable of autonomously executing complex tasks from end to end based on an initial set of user instructions. However, unlike Operator<\/a>, Deep Research is designed for in-depth, extensive workflows, which is why it is branded as a researcher and has capabilities similar to that of a human doing advanced research in various fields. The AI agent can take anywhere from five to 30 minutes to return results depending on the task\u2019s complexity, similar to the extended amount of time it takes a human to complete complex research.<\/p>\n

    \u201cDeep research is built for people who do intensive knowledge work in areas like finance, science, policy, and engineering and need thorough, precise, and reliable research. It can be equally useful for discerning shoppers looking for hyper-personalized recommendations on purchases that typically require careful research, like cars, appliances, and furniture,\u201d OpenAI said in its announcement.<\/p>\n

    \n
    \n
    \n
    \n

    Today we are launching our next agent capable of doing work for you independently\u2014deep research.<\/p>\n

    Give it a prompt and ChatGPT will find, analyze & synthesize hundreds of online sources to create a comprehensive report in tens of minutes vs what would take a human many hours. pic.twitter.com\/03PPi4cdqi<\/a><\/p>\n<\/div>\n

    \u2014 OpenAI (@OpenAI) February 3, 2025<\/a><\/p><\/blockquote>\n<\/div>\n<\/figure>\n

    In OpenAI\u2019s demo, Deep Research was showcased conducting comprehensive market research for a product and identifying the best skis for a user based on their specifications.\u00a0<\/p>\n

    Deep Research is exclusively available to GPT Pro users who pay $200 per month for OpenAI\u2019s premium services. However, OpenAI says both Operator and Deep Research will be available to all tiers of OpenAI subscribers in the future.<\/p>\n

    I see the release of Deep Research as a direct response to the uproar last week over DeepSeek\u2019s entry into the market<\/a>. When DeepSeek made headlines, it triggered doubts about whether United States-based AI companies like OpenAI could justify their sky-high valuations. OpenAI\u2019s move seems like a counterattack to the negative sentiment that Deep Seek cast over U.S. tech companies, a way to shift the conversation back in their favor and remind the world why they remain a dominant force in AI.<\/p>\n

    DeepSeek\u2019s rapid rise and fall: From AI disruptor to global security threat<\/strong><\/p>\n

    Last week, DeepSeek <\/a>took the AI world by storm, making headlines for allegedly matching OpenAI\u2019s performance while spending just $6 million to develop its generative AI (GenAI) model compared to the multi-billions OpenAI has spent. The news rattled the tech industry and raised concerns about U.S.-based AI firms\u2019 valuations and competitiveness.\u00a0<\/p>\n

    But this week? DeepSeek has almost completely disappeared from the conversation. DeepSeek has been banned from government devices in South Korea, Australia, and Taiwan due to national security concerns. Microsoft (NASDAQ: MSFT<\/a>) is investigating whether DeepSeek used OpenAI research and data without authorization, and corporations, as well as government agencies, worry that user data entered into DeepSeek is being funneled directly to China, creating a security risk.<\/p>\n

    All of these elements, along with doubts regarding whether DeepSeek was truly able to create its model for the amount of money and amount of computing infrastructure that they said it used, created the perfect storm for the Chinese competitor to get publicly disparaged and practically wiped from our memories in just one weeks time.<\/p>\n

    \n
    \n
    \n
    \n

    We reproduced DeepSeek R1-Zero in the CountDown game, and it just works <\/p>\n

    Through RL, the 3B base LM develops self-verification and search abilities all on its own <\/p>\n

    You can experience the Ahah moment yourself for < $30
    Code:     https:\/\/t.co\/B2IsN1PrXV<\/a><\/p>\n

    Here’s what we learned ???? pic.twitter.com\/43BVYMmS8X<\/a><\/p>\n<\/div>\n

    \u2014 Jiayi Pan (@jiayi_pirate) January 24, 2025<\/a><\/p><\/blockquote>\n<\/div>\n<\/figure>\n

    On top of that, an AI research team at UC Berkeley was able to replicate DeepSeek\u2019s core capabilities for just $30. While this lends some credibility to DeepSeek\u2019s claim of achieving a low-cost AI breakthrough, it also reinforces the statement made by OpenAI CEO Sam Altman<\/a> that being a copycat is easier than being a pioneer. OpenAI in particular had to spend billions forging the path that companies like DeepSeek can now easily follow.<\/p>\n

    Regardless, I don\u2019t think we will hear much more about DeepSeek in the U.S. The U.S. has a vested interest in making sure U.S.-based AI companies remain dominant forces and are most likely playing a role in suppressing foreign inventions like DeepSeek. I expect that to continue as well, whether it be through direct intervention, trade restrictions, or negative PR. It\u2019s clear that foreign AI firms<\/a>\u2014especially those from China\u2014face significant roadblocks to global expansion set in place by the U.S.<\/p>\n

    Google removes AI weapons ban<\/strong><\/p>\n

    Google (NASDAQ: GOOGL<\/a>) has quietly revised its AI ethics policy, removing<\/a> a key clause that previously prohibited the company from pursuing AI applications for weapons and surveillance. This change represents a significant shift in Google\u2019s stance on AI development.<\/p>\n

    Shortly after the update, Google released a blog post<\/a> justifying its decision, \u201cSince we first published our AI Principles in 2018, the technology has evolved rapidly\u2026 AI has moved from a niche research topic in the lab to a technology that is becoming as pervasive as mobile phones and the internet itself. We believe that companies, governments, and organizations should work together to create AI that protects people, promotes global growth, and supports national security,\u201d said the portion of the announcement that addressed its AI ethics.<\/p>\n

    While some Google employees probably joined the company due to its original commitment to ethical AI<\/a> and are probably now disillusioned by this change, from a business standpoint, the move makes sense.<\/p>\n

    The market for national defense is massive, and government contracts are typically lucrative. The AI sector is notoriously expensive, with major players spending billions on infrastructure, training, and operational costs, while many of them have yet to turn a profit on their big investments. Expanding into the defense sector gives AI firms a financially stable customer base with deep pockets and a clear incentive to invest in and use their AI systems.<\/p>\n

    Beyond financial incentives, there\u2019s another reality: if Google doesn\u2019t build these AI-driven national defense systems, one of its competitors will. The AI industry is clearly at a point where the competition is intense, and each competitors\u2019 product is only better, or worse, by a very narrow margin. If companies choose not to explore key revenue streams like national defense, they risk losing market share to one of their competitors willing to go that route.<\/p>\n

    In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership\u2014allowing it to keep data safe while also guaranteeing the immutability of data.\u00a0<\/em>Check out CoinGeek\u2019s coverage<\/em><\/a>\u00a0on this emerging tech to learn more\u00a0<\/em>why Enterprise blockchain will be the backbone of AI<\/em><\/a>.<\/em><\/p>\n

    Watch: Demonstrating the potential of blockchain\u2019s fusion with AI<\/p>\n