Skip to playerSkip to main content
  • 15 hours ago
Transcript
00:00What is Project Mercury? What is it that OpenAI is working on here?
00:04Good question, Ed. According to our reporting, the reporters in Dubai spoke to sources at OpenAI,
00:09got their hands on some document, and the so-called Project Mercury is effectively OpenAI hiring
00:14ex-investment bankers as contractors, people who can help with the modeling tasks, who can help
00:19the AI with training it on how to write prompts, how to execute models, and how to make sure that
00:26they could get into a position where they can do some of the tasks that are currently performed
00:31by associates and analysts on live deals, such as financial modeling. Their work on PowerPoint
00:36presentations and Excel is something that takes up most of the work hours for a lot of these junior
00:41bankers. It's a lot of repetitive tasks. It's a lot of simple tasks, and there is very little strategic
00:47work there, and perhaps not even a whole lot of analytical work there. So if you could use
00:52technology to replace some of that, the hope is that you could free up some of these junior bankers
00:58to do some of the bigger picture stuff. So it will need an evolution of their skill set,
01:04but also could put them into better use that would not leave them jaded early in their career.
01:11Shree, I know the team over at Liquidity are probably watching the show right now,
01:15worried about their audience as well. This is the origin of the please fix meme, right?
01:21This is investment bankers, junior analysts who spend 80 to 100 hours a week working in Excel
01:28on these types of tasks. You cover Wall Street inside and out. You know these people. Just
01:33explain earlier at the top of the show, I said grunt work. Go deep into the grunt work that OpenAI
01:39wants to displace with custom financial models. I mean, you have to think about when you're working
01:43in a live deal, what really are the requirements? If you are dealing with a client, you want to come up
01:49with material, with presentations, whether it is suggesting potential transactions, coming up
01:55with valuation models, industry comps, looking at the overall landscape out there. That is
02:01effectively a lot of open source research that needs to be done. That is using a lot of your
02:07Excel skills and making sure that you can do the modeling that is inputting numbers again and again,
02:13replacing numbers, putting in new estimates and coming up with figures that could outline various
02:19scenarios for your clients. That is stuff that you could argue that technology can do, can replace you in
02:26doing that and can do it better. And that's where the real promise here comes. But also remember, and this is
02:33something where you spoke about liquidity, but anyone who's followed the space for long knows that when you talk
02:39about junior banker workers, about those 80 to 100 hour work weeks, often that figure is magnified because bankers,
02:45once they finish their work, are waiting for hours on end, waiting for feedback from their managers who then come back
02:52with immediate responses and fixes that you need to do. And there is a revision and you follow that cycle repeatedly
02:58till your manager is satisfied. So perhaps what you also need is an AI solution that can take care of that manager who will
03:05stop focusing on those very little small font level changes and allow you to focus on big picture stuff.
Be the first to comment
Add your comment

Recommended