You don't need me to tell you this... but "they" (ie: executives) don't care. If there's an efficiency ratio to (potentially) be improved - they will take it. Short-term thinking.
"And where we had thought to be alone we shall be with all the world"
It's not, stop being doomers when you don't even understand the tech omg
"If you don't have any enemies in life you have never stood up for anything" - Winston Churchill |
"It's a testament to the sheer belligerence of the profession that people would rather argue about the 'risk-adjusted returns' of using inferior tooth cleaning methods." - kellycriterion
ok relax the guy is just asking, obviously he doesn't understand it, that's why it's framed as a question. if you're so knowledgeable on the topic, care to share some of your wisdom, oh wise one? how can we peasants better understand the tech omg and avoid being doomers?
ok relax the guy is just asking, obviously he doesn't understand it, that's why it's framed as a question. if you're so knowledgeable on the topic, care to share some of your wisdom, oh wise one? how can we peasants better understand the tech omg and avoid being doomers?
I do declare no. The peasants shall remain ignorant, amusingly so.
"If you don't have any enemies in life you have never stood up for anything" - Winston Churchill |
"It's a testament to the sheer belligerence of the profession that people would rather argue about the 'risk-adjusted returns' of using inferior tooth cleaning methods." - kellycriterion
I've been using it as a PE mid level and it's pretty incredible. Still think we need associates though. It can do some things fantastically but can't do everything required to close a deal.
What kind of things are you able to do with it? I generally use my company's internal GPT for organizing notes and helping with slide title options and Gemini/ChatGPT for general quick market research.
I finally bit the bullet and got a Pro subscription. I'm toying around with the excel plug-in over the next few days. I'll update this thread with my thoughts. Or maybe make a new thread. Anthropic pls pay me.
16 comments and only two people are getting ready to try it…take the plunge yourself don’t dawdle or you’ll become obsolete
It’s incredible - better at research if you use it correctly (“pls interview me on areas i may have overlooked and highlight topics or offshoots that should be researched ”)
It can make interactive reports if needed
It can code / manage coding agents to customize things you need done.
Excel isn’t always the best way to store, calculate data, at scale you want things stored in data bases.
It can do data entry and double check it, it can go beyond a paywall to pull info, basically anything a person can do.
It can your own notes and augment it with public data and create a more comprehensive primer.
Give it enough context it can draft memos (or outlines if you prefer more control) that you would have written.
Instead of chasing idiotic lawyers around to redline documents to review markups or whatever, you can do it on a whim with immediate results. Do I trust it to draft docs for me? No. But do I trust legal to draft docs for me, not really either…
Is it perfect? No. Does it make mistakes? Yes. But so does the average analyst/associate. The nice thing is you can spin up other agents based on other models it can talk to to double check. And you can have ten of them working if you want. You can make it work at 3am without feeling bad. It won’t give you attitude or complain about its bonus (although only a matter of time til they start repricing and raising prices)
WS Raider - I’ve only used the Claude excel plugin and Claude Sonnet for drafting memos and charts, but my experience so far doesn’t really match yours. I’m optimistic about the tech in the future but it doesn’t seem like it’s quite there as of right now. Maybe that’ll be different in 6 months to a year idk.
Ok so for context I'm an incoming analyst at Accenture Strategy and I always messed a lot with coding with AI and recently Claude code. Let me tell you that the capabilities are INSANE. I built a full SaaS in 2 months without a SE background. Right now the UI is quite hard for non technical users to use hence why they build Claude Cowork (Claude code but more user friendly). They also launched Claude excel and Claude ppt. I am 100% convinced that in a few years consulting either: Won't need juniors or juniors will have a COMPLETELY different role. New models keep shipping crazy improvements and it's not likely to stop. I think junior lawyers are currently hit the most, banking, PE and IB is still a bit protected by the fact that model don't get excel to a human level yet (close tho) and same for ppt. But in 5 years? Obviously. A lot of people are oblivious to how powerful Codex and Claude code are like for example here is what I built with it to help students and professionals train on live consulting case interviews with AI it's called caseinterviewai (IT'S CRAZY). Hope this helps :)
You said you built an entire SaaS - what do you mean by this? You built a product that either you or other people use regularly? Would you be comfortable selling subscriptions and then taking the flak if there are security issues?
I say this because I see a lot of non-technical people hyping up coding agents. I’m sure they are incredible. But widespread use by non-technical people seems like a recipe for disaster because you don’t know what you don’t know.
To be clear - I don’t say this as a person who is dismissive of AI. I think AI (particularly in domains like coding and math with verifiable answers) is incredible powerful. But from what I understand it is far from perfect. That coupled with the idea that the last 5% is harder than the first 95% gets you a dangerous situation where non-technical people vastly overestimate their ability to use technology outside its intended domain.
This is interesting, I hope my org gives us Claude access (work at an Accenture competitor). That being said, there are weeks I do like 10 hours of work and just spit in everything I need into ChatGPT / Copilot with effective prompts. Did what I needed to do for work today in ~6 hours while taking multiple breaks + watching shows in the background (and it's nearly 40% of what I had to get done this week for a client project). Could have easily required 15+ hours.
Love this take on a morphing economy but I still think many will take that soul-sucking corporate culture.
i.e. Job security, healthcare coverage (much easier to pay far less through your employer rather than being a small business, getting healthcare coverage as an entrepreneur can be easily ~$10K+ a year for yourself + your family), 401K match - droping these and taking a risk can be hard to stomach once you get to a certain point in your career / life.
The only ones I know taking entrepreneurial risks IRL to be yeoman type CEOs all have either nest eggs / safety nets from their well-to-do parents or have already been financially successful in previous ventures. Every year, I hear at least of 1 or 2 more people from my high school graduating class going this route but I know all their parents are well to do and they will be inheriting at least high 6 figures.
This stuff is getting scary. Most jobs will be gone soon. AI automates any worker who applies above-average reasoning to media, text, and audio input. This is 99% of the knowledge economy.
I used to be one of the "AI is a bubble" people, like OzymandiaCRE and [moneymaker03], but I've done a full 180 since discovering tools like Cursor/Claude Code. It's possible now to build a company from the ground up on your own. It's often easier than hiring an additional employee or contractor.
The only thing scary is people falling for techbro marketing talk and delusions. LLMs are not going to eliminate "most jobs" anytime soon. That is just a fundamental misunderstanding of its capabilities. Some theoretical future AI could, sure, and current executives certainly seem to want to eliminate as many jobs as they can out of their own personal greed, but ChatGPT or whatever is not capable of high level functioning.
The one place I think you may have a point are the type of "fake email jobs" that only exist to generate pointless workslop. Think spending a week writing and revising a report that you already know the recipient isn't going to read. But that is less about "AI" capabilities and more how much pointless time-wasting exists in modern corporate America. And no matter who writes the report—Claude or a 23 year old kid named Claude—no one is still going to read it.
Beyond that, it has always been possible to build a company from the ground up on your own.
This stuff is getting scary. Most jobs will be gone soon. AI automates any worker who applies above-average reasoning to media, text, and audio input. This is 99% of the knowledge economy.
I used to be one of the "AI is a bubble" people, like OzymandiaCRE and [moneymaker03], but I've done a full 180 since discovering tools like Cursor/Claude Code. It's possible now to build a company from the ground up on your own. It's often easier than hiring an additional employee or contractor.
The only thing scary is people falling for techbro marketing talk and delusions. LLMs are not going to eliminate "most jobs" anytime soon. That is just a fundamental misunderstanding of its capabilities. Some theoretical future AI could, sure, and current executives certainly seem to want to eliminate as many jobs as they can out of their own personal greed, but ChatGPT or whatever is not capable of high level functioning.
The one place I think you may have a point are the type of "fake email jobs" that only exist to generate pointless workslop. Think spending a week writing and revising a report that you already know the recipient isn't going to read. But that is less about "AI" capabilities and more how much pointless time-wasting exists in modern corporate America. And no matter who writes the report—Claude or a 23 year old kid named Claude—no one is still going to read it.
Beyond that, it has always been possible to build a company from the ground up on your own.
Can’t believe those greedy CEOs want to see the same output created by less input and shift out the production possibility frontier ! I hate them too
Take a random IM for a new deal (even if not real), a bunch of internal materials from a previous deal (model, memos, IC decks etc.) Put them all in a folder. Download and use Claude Code. Set the model to Opus 4.6 Tell it to get up to speed on new deal and do some initial reseach. Then get it to iterate on some internal materials (prepare model, memo etc.) using the same format and template from previous deal.
For a portfolio company, put all the board decks, management accounts, audited financials, models, past IC memos, LP updates etc in a folder. Get claude code to make dashboards, write new LP update memos, strategic reviews etc.
Ea reiciendis et vel impedit quisquam. Ipsum voluptatem consequatur tempore rem provident. Id eum velit aut ex qui nulla excepturi.
Aut perferendis sed dolor vel qui numquam id. Magnam et repellat est eum distinctio eos ducimus. Iusto et minima fuga. Omnis aut eligendi et nesciunt laudantium recusandae. Sit nisi voluptas quidem veritatis aut. Harum minima impedit molestiae quis.
Veritatis ea deserunt ex consequatur voluptatem in quis iure. Inventore fugit velit deleniti rerum et dolores id numquam. Delectus fugiat eligendi non numquam molestiae et temporibus. Aut corporis debitis beatae sit ut quidem. Aliquid ut quia est qui rerum et. Dolorem quia inventore ea soluta nobis minus rem libero.
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)
Cumque expedita laborum est vero. Laboriosam enim nesciunt consectetur explicabo nulla unde tempora. Consequatur excepturi ut voluptas sit voluptatum ipsa. Nam impedit ipsam vero amet.
Sorry, you need to login or sign up in order to vote. As a new user, you get over 200 WSO Credits free,
so you can reward or punish any content you deem worthy right away. See you on the other side!
Has anyone actually tried it? What does it do? Ive only heard of it through the social media hype ecosystem (which may be astroturfed).
That's exactly what I was wondering and wanted to find out on here
If you get rid of junior bankers, who will ever become senior bankers?
That's the whole conundrum. But some teams will definitely get much smaller.
You don't need me to tell you this... but "they" (ie: executives) don't care. If there's an efficiency ratio to (potentially) be improved - they will take it. Short-term thinking.
It's not, stop being doomers when you don't even understand the tech omg
ok relax the guy is just asking, obviously he doesn't understand it, that's why it's framed as a question. if you're so knowledgeable on the topic, care to share some of your wisdom, oh wise one? how can we peasants better understand the tech omg and avoid being doomers?
I do declare no. The peasants shall remain ignorant, amusingly so.
I've been using it as a PE mid level and it's pretty incredible. Still think we need associates though. It can do some things fantastically but can't do everything required to close a deal.
Would you reckon that as they keep improving the tool you may reduce your headcount?
Hopefully not. Always more we could be deploying or helping portcos. Think it just let's us do more.
What kind of things are you able to do with it? I generally use my company's internal GPT for organizing notes and helping with slide title options and Gemini/ChatGPT for general quick market research.
I finally bit the bullet and got a Pro subscription. I'm toying around with the excel plug-in over the next few days. I'll update this thread with my thoughts. Or maybe make a new thread. Anthropic pls pay me.
Thanks, super interested in this. I can't trust the Twitter hype machine.
16 comments and only two people are getting ready to try it…take the plunge yourself don’t dawdle or you’ll become obsolete
It’s incredible - better at research if you use it correctly (“pls interview me on areas i may have overlooked and highlight topics or offshoots that should be researched ”)
It can make interactive reports if needed
It can code / manage coding agents to customize things you need done.
Excel isn’t always the best way to store, calculate data, at scale you want things stored in data bases.
It can do data entry and double check it, it can go beyond a paywall to pull info, basically anything a person can do.
It can your own notes and augment it with public data and create a more comprehensive primer.
Give it enough context it can draft memos (or outlines if you prefer more control) that you would have written.
Instead of chasing idiotic lawyers around to redline documents to review markups or whatever, you can do it on a whim with immediate results. Do I trust it to draft docs for me? No. But do I trust legal to draft docs for me, not really either…
Is it perfect? No. Does it make mistakes? Yes. But so does the average analyst/associate. The nice thing is you can spin up other agents based on other models it can talk to to double check. And you can have ten of them working if you want. You can make it work at 3am without feeling bad. It won’t give you attitude or complain about its bonus (although only a matter of time til they start repricing and raising prices)
157th WSO post about AI replacing junior bankers
WS Raider - I’ve only used the Claude excel plugin and Claude Sonnet for drafting memos and charts, but my experience so far doesn’t really match yours. I’m optimistic about the tech in the future but it doesn’t seem like it’s quite there as of right now. Maybe that’ll be different in 6 months to a year idk.
Ok so for context I'm an incoming analyst at Accenture Strategy and I always messed a lot with coding with AI and recently Claude code. Let me tell you that the capabilities are INSANE. I built a full SaaS in 2 months without a SE background. Right now the UI is quite hard for non technical users to use hence why they build Claude Cowork (Claude code but more user friendly). They also launched Claude excel and Claude ppt. I am 100% convinced that in a few years consulting either: Won't need juniors or juniors will have a COMPLETELY different role. New models keep shipping crazy improvements and it's not likely to stop. I think junior lawyers are currently hit the most, banking, PE and IB is still a bit protected by the fact that model don't get excel to a human level yet (close tho) and same for ppt. But in 5 years? Obviously. A lot of people are oblivious to how powerful Codex and Claude code are like for example here is what I built with it to help students and professionals train on live consulting case interviews with AI it's called caseinterviewai (IT'S CRAZY). Hope this helps :)
You said you built an entire SaaS - what do you mean by this? You built a product that either you or other people use regularly? Would you be comfortable selling subscriptions and then taking the flak if there are security issues?
I say this because I see a lot of non-technical people hyping up coding agents. I’m sure they are incredible. But widespread use by non-technical people seems like a recipe for disaster because you don’t know what you don’t know.
To be clear - I don’t say this as a person who is dismissive of AI. I think AI (particularly in domains like coding and math with verifiable answers) is incredible powerful. But from what I understand it is far from perfect. That coupled with the idea that the last 5% is harder than the first 95% gets you a dangerous situation where non-technical people vastly overestimate their ability to use technology outside its intended domain.
This is interesting, I hope my org gives us Claude access (work at an Accenture competitor). That being said, there are weeks I do like 10 hours of work and just spit in everything I need into ChatGPT / Copilot with effective prompts. Did what I needed to do for work today in ~6 hours while taking multiple breaks + watching shows in the background (and it's nearly 40% of what I had to get done this week for a client project). Could have easily required 15+ hours.
Love this take on a morphing economy but I still think many will take that soul-sucking corporate culture.
i.e. Job security, healthcare coverage (much easier to pay far less through your employer rather than being a small business, getting healthcare coverage as an entrepreneur can be easily ~$10K+ a year for yourself + your family), 401K match - droping these and taking a risk can be hard to stomach once you get to a certain point in your career / life.
The only ones I know taking entrepreneurial risks IRL to be yeoman type CEOs all have either nest eggs / safety nets from their well-to-do parents or have already been financially successful in previous ventures. Every year, I hear at least of 1 or 2 more people from my high school graduating class going this route but I know all their parents are well to do and they will be inheriting at least high 6 figures.
The only thing scary is people falling for techbro marketing talk and delusions. LLMs are not going to eliminate "most jobs" anytime soon. That is just a fundamental misunderstanding of its capabilities. Some theoretical future AI could, sure, and current executives certainly seem to want to eliminate as many jobs as they can out of their own personal greed, but ChatGPT or whatever is not capable of high level functioning.
The one place I think you may have a point are the type of "fake email jobs" that only exist to generate pointless workslop. Think spending a week writing and revising a report that you already know the recipient isn't going to read. But that is less about "AI" capabilities and more how much pointless time-wasting exists in modern corporate America. And no matter who writes the report—Claude or a 23 year old kid named Claude—no one is still going to read it.
Beyond that, it has always been possible to build a company from the ground up on your own.
Can’t believe those greedy CEOs want to see the same output created by less input and shift out the production possibility frontier ! I hate them too
Try it yourself:
Ea reiciendis et vel impedit quisquam. Ipsum voluptatem consequatur tempore rem provident. Id eum velit aut ex qui nulla excepturi.
Aut perferendis sed dolor vel qui numquam id. Magnam et repellat est eum distinctio eos ducimus. Iusto et minima fuga. Omnis aut eligendi et nesciunt laudantium recusandae. Sit nisi voluptas quidem veritatis aut. Harum minima impedit molestiae quis.
Veritatis ea deserunt ex consequatur voluptatem in quis iure. Inventore fugit velit deleniti rerum et dolores id numquam. Delectus fugiat eligendi non numquam molestiae et temporibus. Aut corporis debitis beatae sit ut quidem. Aliquid ut quia est qui rerum et. Dolorem quia inventore ea soluta nobis minus rem libero.
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...
Cumque expedita laborum est vero. Laboriosam enim nesciunt consectetur explicabo nulla unde tempora. Consequatur excepturi ut voluptas sit voluptatum ipsa. Nam impedit ipsam vero amet.