Goldman Sachs unveils AI assistant as technology takes over banking

goldman sachs Chief Information Officer Marco Argenti said the company is rolling out generative AI assistants for bankers, traders and asset managers that will eventually take on the characteristics of experienced Goldman employees. This is the first step in the evolution of the program.

In an exclusive interview with CNBC, Argenti said the bank has released a program called GS AI Assistant to about 10,000 employees and will make it available to all knowledge workers in the company this year. He said that is his goal. Initially, it’s useful for tasks like summarizing and proofreading emails, and translating code from one language to another.

“Think of all the tasks you might want to complete in terms of different use cases for all the out-of-the-box professions,” Argenti said. Goldman Assistant is a “very simple interface that gives you access to the latest and greatest models.”

Goldman’s move means: JP Morgan Chase and morgan stanleythe world’s top three investment banks are actively releasing generative AI tools for their employees, which has been a remarkable development since ChatGPT took off about two years ago.

Wall Street has embraced generative artificial intelligence faster than any other disruptive technology in recent years, experts say, because large-scale language models are extremely effective at replicating aspects of human cognition. It is said that it is because it is superior.

Currently it can respond to queries, compose emails, summarize long documents, etc., but future versions will include so-called “agent” functionality, i.e. multi-step processing with little human intervention. The ability to perform tasks is expected.

In an interview with CNBC, Argenti talked about his vision for artificial intelligence at the company. Amazon In 2019, we repeatedly likened the AI ​​program to new employees absorbing Goldman’s culture over the next few years.

Initially, the tool will primarily generate answers based on Goldman data fed into the AI ​​model from OpenAI’s ChatGPT. Google with gemini meta Argenti said llamas vary depending on the job. He added that the bank is also considering models from companies such as Anthropic, Mistral and Cohere.

“The AI ​​assistant will really be like talking to another GS employee,” Argenti said.

“As the work progresses, the second step is this agentic behavior kicks in, where you say, “I’m completing a task on behalf of a Goldman employee, so I need to take a series of steps.” ’,” he said. Said. “From there, the models will start not only talking like Goldman employees, but acting like Goldman employees.”

This helps explain why companies have begun to prohibit their employees from using ChatGPT at work and instead create their own platforms to utilize the technology. This allows companies to not only keep their information safe, but also build AI platforms that increasingly resemble the best examples of their own employees.

“It’s very important that AI has a very specific identity that reflects the company’s philosophy, values, knowledge and mindset,” Argenti said.

In practice, this is similar to how experienced Goldman employees know to use multiple data sources to double-check their work or use specific algorithms for calculations. This means that AI will absorb those lessons, he said.

But Argenti said he’s most excited about the prospect of what will happen in perhaps three to five years, as AI models increasingly blur the lines between human and machine thinking. I am.

At this stage of Goldman’s AI, he said, the model will “actually reason more and get closer to the way Goldman employees think.”

So instead of being handed a runbook, a technology-industry term meaning a set of step-by-step instructions for completing a task or responding to an incident, the AI ​​is You will be able to create detailed plans in a way that I will,” Argenti said.

That forward-looking outlook is helping Wall Street employees train on technology that could make some roles obsolete while enhancing others and creating entirely new ones. This fact may send a new wave of anxiety among the workforce.

Like Goldman, other major investment banks are aiming to offer generative AI tools to all employees in the coming months.

More than 200,000 employees currently have access to the company’s generative AI tools, according to a person familiar with JPMorgan Bank, who requested anonymity to discuss internal matters. The bank announced in October that about 40,000 Morgan Stanley employees had access to the service as of the end of last year.

Finance and technology is considered one of the industries where employees are most likely to be disrupted due to generative AI, with companies potentially able to generate billions of dollars in additional profits. Meta CEO Mark Zuckerberg told podcaster Joe Rogan earlier this month that his company’s AI will be able to write code like an intermediate-level software engineer this year. He said he was deaf.

Global investment banks could cut up to 200,000 jobs over the next three to five years as they adopt AI, according to a report from Bloomberg’s research division. The report, based on a survey of technology executives at major banks, said support and operations roles, known as back and middle offices, are most at risk.

But Goldman’s official position is that AI will enable employees to do more, not necessarily reduce the need for humans.

“The importance of having a phenomenal human workforce is actually going to be amplified even more,” Argenti said.

“In my opinion, at the end of the day, it’s a human issue,” he says. “People are going to make a difference, because they are the ones who are actually going to evolve AI, educate AI, empower AI, and take action.”