These "Magical Terms": Data Science, Machine Learning, AI... in IB
Hi All.
...Data Science...
...Neural Networks...
...Machine Learning...
...Artificial Intelligence...
=> Seem like sort of Magical Terms (almost like "landing on Mars"... or buying Twitter) but I guess they are used more & more in the Financing World.
I wonder where have you seen practical applications of Data, Data Science & Machine Learning in Finance & Investment Banking?
What do you think will be the demand for Data Scientists & Machine Learning experts in IB in the coming years and how can they impact Investment Banking and the wider Finance world?
What's the Future?
Best,
Eric
Del
I'm a black and white guy and what I see here, is the person who prevents people from a discussion by posting every single time:
"Stop talking - search for it because it was answered many times."
(example is the link in his reply)
And then he provides an example of a "discussion" with ONE reply.
His.
And that's the point when I straight away think => what a piece of Monkey S*it.
It's even more interesting than when you search on WSO for the terms like "Data Science" or "Machine Learning" => then the only thread that has this term exactly in => is this one.
============
On the other hand:
Are you posting the same when people have interest in comps in IB and ask "what $$$ to expect in A, B or C" for a 1,000th time?
"Stop talking about IB comps. Search for it."
Are you?
=============
I need to ask you:
Stop being an aggressive, conversation-stopping piece of Monkey S*it.
Let people be and have a conversation that is longer than a single reply because the technology for finance is moving, people are interested and they want to learn.
Or simpler, served on BLACK AND WHITE plate for you:
if you have nothing to say => kindly stop talking.
And let others contribute.
All right?
All the Best,
Eric
* EDIT * If, like 11 other users (at the time of this edit), 1) you find this comment at least slightly funny/helpful & 2) you support an open conversation on WSO - then please also give a suitable "reward" to a person who "caused" this reply (and whose original comment is now edited to "del" above this reply... probably due to hmmm... "interesting" banana-to-crap "reward ratio")
...because people like to contribute. Not to be stopped.
ML and AI are massively overblown by people who don't really understand what they are.
There are two big problems with ML/AI in a finance context.
First, is the black box nature of ML/AI models. By their nature, it is extremely difficult to pinpoint what caused a ML/AI model to make a decision in real time. This is problematic in the financial world. ML/AI models are at their core, sophisticated statistical models. They take in data, learn patterns or rules, and output decisions or probabilities. With as much bad or misleading data as there is out there, this can quickly become problematic. And bad decisions or output are extremely difficult to catch before its too late, given how opaque the models are. This might be ok for hedge funds or prop funds who are willing to live on knifes edge, but its a tougher sell for singular large transactions like you'd find in IBD, PE, some sell side S&T etc. We're getting into some real philosophical, Isaac Asimov/iRobot bullshit here, but I tend to think there is a pretty significant unclose-able gap between organic living intelligence and artificial intelligence. That parts just my opinion though.
Second problem is regulation. I worked in credit risk at a commercial/consumer bank at one point. Part of my job was to find the bank new ways to mitigate losses through better credit models. ML/AI models are a huge regulatory no-no. The CFPB/OCC/Feds will be down your throat faster than you can say "Elon". We had to be able to demonstrate to regulators why every credit decision we made was not discriminatory in terms of protected classes (race, religion, sexuality, etc.). Opaque ML/AI models are not compatible with that framework. This is more anecdotal, but ML/AI models in credit have a tendency to get very racist very very quickly when you train them on default data. So uh... Don't do that I guess. Anyways, point being, if I cant approve/decline John Doe of Springfield Iowa for a credit card with a $3,000 limit using AI/ML, how do you think our friends in the federal government will feel about deciding key aspects of M&A transactions or securities issuance primarily with ML/AI?
I went to school for statistics and related topics. I have a deep respect for Stats/ML/AI and don't mean to shit on them unnecessarily. They are incredibly useful tools. Just saying they make a lot more sense in less regulated industries (targeted marketing for example is a great usecase) where the cost of mistakes is much lower or the tolerance of risk is much higher. Is it possible that banks will start hiring exclusively data science PhDs to run IBD teams? Sure. Do I think its likely? No. Do I think it will happen during either of our careers? Lol.
Del
don't think he meant black and white in that way lol
Deserunt sit neque qui et provident aperiam et vitae. Asperiores rem neque magnam dolorum. Occaecati alias qui facilis. Qui repellat aut quasi voluptates quia. Commodi ad ipsum quisquam aperiam ab eum in. Quisquam quaerat enim impedit sit maiores et voluptatem. Suscipit est libero qui sit quo et cumque.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Enim est odit eaque tempore non. Tempore velit nesciunt aut facilis esse voluptas ab. Perspiciatis quo omnis vero est.