OpenAI Looks to Replace Banking Grunt Work

Key Concepts

Project Mercury: OpenAI's initiative to leverage AI for tasks currently performed by junior investment bankers.
Investment Bankers (Associates and Analysts): Junior roles in investment banking responsible for tasks like financial modeling, presentation creation, and data analysis.
Grunt Work: Repetitive, time-consuming, and less strategic tasks in investment banking, such as data input, basic modeling, and presentation formatting.
Financial Modeling: The process of creating mathematical models to forecast a business's financial performance, often involving complex Excel spreadsheets.
Prompt Engineering: The skill of crafting effective prompts to guide AI models to produce desired outputs.
AI-driven Automation: The use of artificial intelligence to automate tasks previously done by humans.
Managerial Feedback Loop: The iterative process of junior bankers submitting work, receiving feedback from managers, and making revisions.

Project Mercury: OpenAI's Initiative to Automate Junior Banking Tasks

OpenAI is reportedly working on a project, codenamed "Project Mercury," aimed at automating tasks traditionally handled by junior investment bankers, such as associates and analysts. This initiative involves hiring investment bankers as contractors to assist in developing and training AI models. The goal is to equip these models with the capability to perform tasks like financial modeling, creating PowerPoint presentations, and working with Excel spreadsheets, which currently consume a significant portion of junior bankers' work hours.

The Nature of Junior Banker "Grunt Work"

The transcript highlights that the work performed by junior bankers often consists of repetitive, simple tasks with limited strategic or analytical depth. These tasks include:

Financial Modeling: This involves extensive use of Excel to input and manipulate numbers, create valuation models, and outline various financial scenarios. It requires continuous updating of estimates and figures.
Presentation Creation: Developing materials and presentations for clients, which can include suggesting potential transactions and analyzing industry landscapes.
Data Input and Manipulation: Repeatedly entering and replacing numbers, a core component of financial modeling and analysis.

These tasks are often described as "grunt work," contributing to the demanding 80-100 hour workweeks common in investment banking.

The Promise of AI in Investment Banking

The core argument for Project Mercury is that technology, specifically AI, can perform these repetitive tasks more efficiently and potentially better than humans. The hope is that by automating this "grunt work," junior bankers can be freed up to focus on more strategic, higher-level activities. This would not only improve their utilization but also potentially prevent career burnout and disillusionment.

Addressing the Managerial Feedback Loop

Beyond task automation, the transcript suggests that AI could also address the inefficiencies of the managerial feedback loop. Junior bankers often spend hours waiting for feedback from their managers, only to receive immediate requests for minor revisions. This iterative cycle of submission, feedback, and revision can be time-consuming and frustrating. Project Mercury could potentially offer solutions that streamline this process, allowing managers to focus on substantive issues rather than "very little small font level changes."

Real-World Implications and Industry Impact

The discussion draws a parallel to the "please fix" meme, which originates from the experiences of junior analysts facing constant revisions. The initiative by OpenAI, if successful, could significantly impact the investment banking industry, particularly roles focused on data-intensive and repetitive tasks. Companies like Liquidity, which likely cater to this segment of the financial industry, are noted as potentially observing this development with concern regarding their own services.

Technical Aspects and Methodologies

While specific technical details are not extensively elaborated, the process involves:

Hiring Investment Bankers as Contractors: To provide domain expertise and practical insights into the tasks that need automation.
AI Model Training: Teaching AI models how to perform financial modeling, prompt engineering, and execute complex tasks.
Prompt Engineering: Developing effective prompts to guide the AI in generating accurate and relevant outputs for financial analysis and presentations.
Model Execution: Ensuring the AI models can reliably perform the identified tasks.

Key Arguments and Supporting Evidence

Argument: Junior banker tasks are largely repetitive and can be automated by technology.
- Evidence: Description of financial modeling, presentation creation, and data input as time-consuming, simple, and lacking strategic depth. The 80-100 hour workweeks are cited as a consequence.
Argument: AI can perform these tasks more efficiently and effectively.
- Evidence: The inherent capability of technology to handle repetitive computations and data manipulation.
Argument: Automating grunt work will free up junior bankers for higher-value activities.
- Evidence: The potential for junior bankers to focus on "bigger picture stuff" and strategic thinking.
Argument: AI can improve the efficiency of the managerial feedback loop.
- Evidence: The description of the iterative and often time-consuming process of revisions based on managerial feedback.

Conclusion

Project Mercury represents OpenAI's ambitious effort to disrupt the traditional workflows of junior investment banking by leveraging AI. By automating the "grunt work" of financial modeling, presentations, and data analysis, the initiative aims to enhance efficiency, improve the utilization of junior talent, and potentially alleviate the intense work hours and burnout experienced in the industry. The success of this project could lead to a significant evolution in the skill sets required for junior bankers and reshape the operational landscape of investment banking.