The Goldman Sachs logo displayed on a smartphone.
Budrul Chukrut | Sopa Images | Lightrocket | Getty Images
Goldman Sachs is experimenting with generative AI tools internally to help its developers automatically generate and test code, the company’s chief information officer told CNBC.
Marco Argenti, who joined Goldman as a partner from Amazon in 2019, said Tuesday that the firm’s software engineers have been using the technology to automatically generate lines of code.
It is currently in a “proof of concept” stage and not yet ready for production, he added.
“Developers are already using some of the assisted coding technology,” Argenti told CNBC’s Arjun Kharpal at the Goldman Sachs technology symposium on Tuesday.
Generative AI refers to a group of products that produce human-like text or images in response to written prompts from users.
Among the most popular examples is ChatGPT, an AI chatbot developed by Sam Altman’s OpenAI. Other competing products include Google’s Bard and Stable Diffusion, an AI-based image generator created by startup Stability AI.
Goldmans’ interest in generative AI products comes despite pushback from some banking giants on the use of ChatGPT internally. JPMorgan, Goldman Sachs, Citigroup and Bank of America have all reportedly restricted staff from using the software.
Argenti declined to identify the generative AI products that the company has been using. He also did not specify which bank division the tech is being used in.
ChatGPT and products like it could potentially radically shake up the world of financial services. AI could take the steering wheel on which investment decisions to make, for example, or automate many customer service functions.
“It’s still very early,” Argenti conceded, although he compared the development of generative AI to “the beginning of the internet.”
“You wouldn’t put immediately all your most important workloads there, but the imperative is to really to try to understand the potential,” he added.
Goldman has invested…
Read the full article here
