OpenAI Is Building a Banker

The classic description of a successful investment banking analyst is “detail-oriented.” When you arrive at an investment bank fresh out of college, you will be asked to prepare materials for client meetings and to build financial models in Excel. How can you demonstrate that you are good at the job? You just got there; you are unlikely to have any brilliant insights into the client’s needs. Pretty much you’re going to put together the materials your bosses tell you to, using pages from previous client pitches. You can do this sloppily in a way that embarrasses your bosses: Every investment banker has been in a client meeting where the pitchbook has the wrong logos because the analyst copied from a precedent without updating it. No one likes that! Or you can do it sloppily in a way that doesn’t embarrass your bosses, but that causes them to seethe with secret rage: Every investment banker has also seen a vice president get furious about a table with dollar signs aligned in two different ways. This rage is not substantively justified, but the VP is worried that if the dollar signs are misaligned then there might be a worse error somewhere in the book. Or you can just do it perfectly, and then the VP will like you.

Meanwhile, if you are building big financial models with a lot of moving pieces, they will take you hours to build, but they will also take a long time for your bosses to check. If you demonstrate that you don’t need checking — if you build your models right the first time, if you follow best practices in modeling and formatting, if everything looks clean, and also if your pitchbooks are error-free — then your bosses will trust you and appreciate your work. If you make them nervous — if your models have obvious mistakes, or if your pitchbook formatting is bad — then they will have to spend a lot of time checking your work and will not appreciate that.