China’s AI Startup Deepseek has made a headline with the latest AI model, and the release of the V3 model last week was just a hot topic. However, what we are really turning is Deepseek-R1, the company’s latest probation model compared to O1 of OPENAI by providing more cost-effective alternatives as well as abilities. DeepSeek has not decelerated. Now we are paying attention to Wall Street.
Goldman Sachs are interested and that’s a big deal
Questions about everyone’s mind now: How fast is DeepSeek’s open source AI going into major business? Especially when you look at Fortune 500 companies, the answer is becoming more clear. The assumption that a major US company to avoid Deepseek is outdated just because it is based in China is proven to be outdated.
According to a new article by Jon Victor, a reporter who covers the Enterprise software and AI, Goldman Sachs is currently investigating the possibility of using DeepSeek.
Goldman’s AI: What has changed?
Marco Argenty, the highest information manager of Goldman Sachs, has confirmed that his team is considering DeepSeek, but adoption needs to be a security review. This is a standard precautionary measure to prevent potential vulnerabilities such as quick injection. This is a problem that must be considered when integrating AI into a workflow, regardless of where the model is.
“Marco Algenty, the highest information manager of Goldman Sachs, said his team was also interested in using Deepseek, but the model needs to be screened. Probably a quick injection. This is to prevent hidden data, regardless of whether the model is Chinese, but is a security concern for companies that enhance the application. Victor is writing.
If Goldman is interested, who is next?
Goldman’s consideration for Deepseek is a signal for larger things: Open source AI models are no longer catching up. They are now serious candidates.
Investment bank approaches to AI reflect a greater trend in corporate technology recruitment. Goldman, especially in software development, has already used AI to improve productivity. Since the introduction of Microsoft’s GitHub Copilot to more than 10,000 engineers in June, the company has increased average coding efficiency by 10 %. ARGENTI equally equated it with adding 1,000 engineers to the staff.
And they have not stopped there. Goldman tests a more advanced AI coding tool, but Argenti was not detailed.
Beyond coding: AI for bankers and traders
It is not only engineers who are looking at improvements in AI -led. Goldman expanded internal AI chatbots for bankers and traders, created client presentation slides, and supported their insights from long reports. Employees can select Meta (LLAMA), Google, and Openai models to select the best fit for each task.
The latest model of DeepSeek has revealed that financial institutions are as high as high -tech companies. However, Argenti has no plans to invest in AI model training from zero. Also, GOLDMAN is not trying to build a unique model.
“We are not the business of the Foundation model,” he said.
Financial AI: Hype -hype
Some industry leaders believe that AI is wise enough to ultimately lead to investment decisions and can revolutionize portfolio management. Turing Jonathan Sidalus suggests that AI models can develop powerful inference skills to predict their assets.
Argenti is not sold in that vision.
He told the information that the AI -drive -type highest investment officer was theoretically wonderful, but there are still restrictions in the real world. At the end of the day, he believes that these models are as good as the data they have trained, and they don’t care about magic.
For now, human judgment still has the final decision.
Everyday role of AI in Goldman
Other than work, Argenti depends on AI tools such as Chatgpt and Google Gemini, maintaining AI research in the latest state and supporting philanthropy. He used AI to build a Seattle -based rock band Element47 website.
“For the web development … incredibly, it’s incredibly efficient,” Argenti told the information.
Open source AI proves that it is not just an alternative, but a competitive force in the industry, as financial giants such as Goldman Sachs are exploring DeepSeek.
On the other hand, Openai’s Sam Altman, the main competitors of Deepseek, broke silence on Wednesday and acknowledged the progress of AI with praise. In the X post, Altman explained that DeepSeek’s R1 model was “especially impressive about what can be provided for the price.”
Altoman is not alone. NVIDIA also praised DeepSeek’s progress when creating a major language model (LLM) with cheaper and more energy efficiency. In a statement to Bloomberg, the chip maker called it “excellent AI progress” and emphasized the compliance of the US technology export control.
Source link
