Bankers become teachers for AI. Former banking brains, paid by the hour to teach robots their trades
More than 100 former investment bankers from some of the world's biggest banks are training artificial intelligence on how to build financial models in a bid to replace the hours of work done by junior bankers in the industry.
100 former bankers teach Artificial Intelligence banking operations. Archive photo
The group, which includes former employees of JPMorgan Chase & Co., Morgan Stanley and Goldman Sachs Group Inc., is part of a secret project within OpenAI codenamed Mercury.
Participants are paid $150 an hour to write materials and build financial models for a wide range of deals, including restructurings and initial public offerings, according to insiders cited by Bloomberg. The company has also given contractors early access to the artificial intelligence it is creating, which it aims to replace entry-level jobs in investment banks.
The project underscores Sam Altman's OpenAI's urgency to make its powerful AI technology more useful to companies in a wide range of industries, from consulting to finance, legal and technology. Although it hit a $500 billion valuation earlier this month, the world's largest startup has yet to turn a profit.
An OpenAI spokesperson said the firm is working with a number of experts “to improve and evaluate the capability of our models in different areas. Experts are recruited, managed and compensated by third party providers”.
Investment banking analysts typically spend more than 80 hours a week at their desks when working on real transactions, building detailed models in Microsoft Corp.'s Excel program. for both mergers and leveraged buyouts. They often face a steady stream of requests from superiors to make changes to PowerPoint slide presentations, and then changes to those changes—a culture that spawned the “pls fix” Wall Street meme.
Already, a lot of startups are trying to step in and equip banks with artificial intelligence that can help in all these situations. While analysts have long complained about this chore, the rise of artificial intelligence is now raising concerns about job security.
The application process for Project Mercury involves almost no human interaction, according to a person familiar with the situation, who asked not to be named to discuss non-public information.
The first step is a roughly 20-minute interview with an AI-powered chatbot that asks questions based on the candidate's resume. The second phase tests candidates' knowledge of financial statements. The final stage is a modeling test.
“The job is flexible and contractors have to submit one model per week“, the person said. The requirements include writing instructions in simple terms, then running the model. Participants get feedback from a reviewer and are expected to fix any problems before their work is eventually integrated into OpenAI systems, the person said.
So far, Project Mercury has attracted participants who previously worked at various Wall Street headquarters, including Brookfield Corp. , Mubadala Investment Co. , Evercore Inc. and KKR & Co. , the documents show. Also participating in this effort are several current MBA candidates from Harvard University and the Massachusetts Institute of Technology.
Participants are asked to create their models in Excel and also adhere to industry norms for formatting models, including areas such as margin sizes and italic percentages.
