Yo, Xponents.

[Unmutual]

You see this NYT article about the future of investment bank analysts?

The Worst Part of a Wall Street Career May Be Coming to an End

Artificial intelligence tools can replace much of Wall Street’s entry-level white-collar work, raising tough questions about the future of finance.

www.nytimes.com

 

[no gas]

I’ve had enough of analysts who are right less than half the time.

 

[Unmutual]

Yeah, but who's gonna order the Chinese takeout at midnight?

 

[Electrified]

Let's analyze how to get rid of the analysts.

 

[pjdcorp]

Let's analyze how to get rid of the analysts.

After we "kill all the lawyers" ??

 

[xponents]

You see this NYT article about the future of investment bank analysts?

The Worst Part of a Wall Street Career May Be Coming to an End

Artificial intelligence tools can replace much of Wall Street’s entry-level white-collar work, raising tough questions about the future of finance.

www.nytimes.com

I must admit I almost had a heart attack upon reading the title of this post with my name in it... I was like, "Oh no.. what did I do now?"

But yes, I did read it! I thought the article was insightful, and I agree that AI will be used as a tool to help analysts, not to replace them. As previously mentioned, I myself am looking to see where AI could or already is being used within investment banking. I already use AI a lot for financial models, but only to automate the "waste" work I would needed to have done such as entering the numbers, researching data (tip: use Gemini for this as base ChatGPT does not have internet access), etc. As the article states, I basically use it to do things that are ordinarily a time-sucker. The time it takes me to do a simple three statement model now takes about 2 days (sounds long, but I have my school work to do too) instead of a week or so prior to AI. I do NOT however trust it to make assumptions for me and forecast for me (other than entering the numbers I give it manually). I'm sure it can do it well most of the time, but I'd rather be safe than sorry with the crap some AI models put out when asked to "think." I'm still doing the actual work, but AI is just automating the process a bit (this is safer to me than AI doing the whole thing).

However, I don't think AI will ever be able to replace humans in this field (sorry @no gas and @Electrified !). There needs to be SOMEBODY to promote to the higher levels, after all. As you go higher and higher (starting from being an associate), my understanding is that the work strays away from slaving away at models and powerpoints for your Associates and VP's and focuses more on communication. Like it or not, AI will never be able to replace that.

(Part one of this post... too long to fit in one message)

 

[Daniel2022AT]

There was a study done at MIT in the late 1980’s about the effect of technology on work life. The image from cartoons in the 1960’s like the Jetsons (remember anyone?) is that the machines do all the work and humans just sit around with a life of leisure. Kurt Vonnegut wrote a great novel on this topic, “Player Piano”. However, the MIT study showed in reality the opposite effect. With more tools and capability, technology just ups the bar on the level of analysis that can be done so it becomes expected. People were working harder and longer as a result. We are so far short of figuring out and optimizing anything and everything there is much to be done, be it structuring a balance sheet, inventing new drugs, or feeding populations. Nobody knows the impact of AI yet. Until every human on earth lives a comfortable full optimized life, there is much work yet to be done.

 

[borski]

There was a study done at MIT in the late 1980’s about the effect of technology on work life. The image from cartoons in the 1960’s like the Jetsons (remember anyone?) is that the machines do all the work and humans just sit around with a life of leisure. Kurt Vonnegut wrote a great novel on this topic, “Player Piano”. However, the MIT study showed in reality the opposite effect. With more tools and capability, technology just ups the bar on the level of analysis that can be done so it becomes expected. People were working harder and longer as a result. We are so far short of figuring out and optimizing anything and everything there is much to be done, be it structuring a balance sheet, inventing new drugs, or feeding populations. Nobody knows the impact of AI yet. Until every human on earth lives a comfortable full optimized life, there is much work yet to be done.

Completely correct. And that’s why, for example, understanding actual computer science and engineering from first principles is far, far more important than “learning Java.”

 

[xponents]

Completely correct. And that’s why, for example, understanding actual computer science and engineering from first principles is far, far more important than “learning Java.”

I must say I agree with both of you, but wouldn't you guys say that this (my) implementation of AI would shorten times instead of lengthening them as @Daniel2022AT stated, since no extra analysis is being added? Isn't it just automating the "waste work?" Obviously technology will be ever-improving, and therefore new functions will be added, but when it's tied to something like a financial model, what else could possibly be added that would be useful to these firms? Not trying to prove my point, just genuinely curious. Do note that I'm talking about investment banking specifically, something like quant HFs are a totally different story (two sigma, etc).

Also, for @borski , can’t you do both at the same time?

 

[maractwin]

can’t you do both at the same time?

Yes, and the best programmers do. There are a few people in this world who only deal in theory but not implementation, a lot of people who churn out code without understanding what really happens, and then the good ones who work in a high level language, but understand what that turns into under the covers. I've had to show a few junior programmers what their correct, but inefficient code turns into. Sure, modern compilers include some amount of source optimization, but that doesn't always work. Though AI may improve that.

 

[borski]

Obviously technology will be ever-improving, and therefore new functions will be added, but when it's tied to something like a financial model, what else could possibly be added that would be useful to these firms?

I’ve said it before and I’ll say it again: you’re not thinking big enough.

Imagine the future. Not the iterative future, but the ideal future. Skip a few steps.

Then figure out how to get there.

That’s how you innovate. Not by iterating on an existing tool.

If anything, async communication is actually the thing AI would be perfect for.

Also, for @borski , can’t you do both at the same time?

Not only can you, but you should. But the point is that learning the theory and first principles means you can pickup basically any language very quickly, instead of starting from scratch every time. Adapting to change is easier if you understand how things works under the good.

 

[Bobby]

There was a study done at MIT in the late 1980’s about the effect of technology on work life. The image from cartoons in the 1960’s like the Jetsons (remember anyone?) is that the machines do all the work and humans just sit around with a life of leisure. Kurt Vonnegut wrote a great novel on this topic, “Player Piano”. However, the MIT study showed in reality the opposite effect. With more tools and capability, technology just ups the bar on the level of analysis that can be done so it becomes expected. People were working harder and longer as a result. We are so far short of figuring out and optimizing anything and everything there is much to be done, be it structuring a balance sheet, inventing new drugs, or feeding populations. Nobody knows the impact of AI yet. Until every human on earth lives a comfortable full optimized life, there is much work yet to be done.

MIT? Is that a community college? @borski

 

[borski]

MIT? Is that a community college? @borski

Massachusetts Institute of Trees

They grow arborists