Freaking out about AI

i still think PE is more than the aggregation of dozens of tasks that UNILATERALLY can be done by AI

but ultimately who knows if you're talking about a few years out. things are moving so fast that most AI experts can't even predict 1 year out. let's see how gamechanging the next versions are, aka chatgpt5 and the like.

good news is, in a few years you won't be a PE junior anymore. this is more a cautionary tale for current ugrad kids

With a single prompt you can build a DCF from scratch! Hmm okay or we can just used the saved template excel file that we’ve had for ten years. There’s nothing special about a model. 

There are going to be lots of efficiency gains but nothing revolutionary. Building an IC memo faster is an efficiency gain. Mapping a market faster is an efficiency gain. Etc. Etc. Hopefully this takes some of the drudgery out of the job. 

I don't fully understand the intricacies of AI/AGI but having used some of the more mainstream tools and tested other specific finance tools, we're still ways away from juniors being redundant. 

1. investing is a forward looking job, and most of these tools produce results based on tons of backtesting. Backtesting can maybe point out some correlation, but isn't always accurate. No LP at your typical MF or UMM or even MM fund is going to pour in millions of dollars if you simply rely on AI tools to invest.. they can go to quant funds for that. People need diversification.
2. in order to get the best response from AI tools, you still have to feed it the right questions + right data, go through iterations of asking questions until you get something acceptable. I've used these tools, and they're excellent in summarizing historical data and providing general industry or business research, but anything beyond that is severely lacking, spits out false information, needs tons of repeated refining of the questions asked. I am sure though that will all the smart people in silicon valley, it's only a matter of time before we get super accurate products.
3. I am obviously biased when I say this but for the most part, it's a productivity enhancing tool, not a replacement. Obviously certain industry and jobs will be made redundant, but certain jobs in finance like yours are a bit far removed from it so far. You also have to realize that you cannot simply have hundreds of thousands or millions of jobs taken away... you will not have a properly functioning society if that happens. The backlash will be extreme. So someone will have to step in, regulate and manage the use cases. If the Dems win in 2028, we'll likely see that to an extreme.
4. you're being rage baited by AI doomsayers.. I know exactly the type of people who do this. I have friends in tech who do this too. Someone on my linkedin who is a principal at a well known VC fund keeps posting AI defeating humans type of crap every single day... guy is an oversmart moron who will post biased content (even where the use case or result of AI is much worse than an actual human) that makes it look like he is investing in the next nvidia every single time. Obviously he is incentivized to make such posts.. it's the tech bro equivalent of the finance bros acting like they're working on facebook's buyout when in reality they're doing LMM HAVC buyouts. Simply ignore it or block that content.

meh, i'm sure some milk delivery man used to make the same complaint with the advent of refrigeration or urine taste tester about new gunpower formula. 

don't need to worry about AI taking over PE because that won't happen for a while, but PE is just no longer the desired best way to make money for sure, like someone who worked at figma for the past 4 years made more money than say a principal at kkr for the same time being, for 60% of the stress and WLB, and there are more unicorns like figma than seats at a kkr equivalent.

I wouldn't worry about it. Just learn to leverage the latest AI tools to get better / faster at your job and you'll be fine. 

High level take is that PE is a zero sum / adversarial industry, so any AI tools that are adopted by one fund need to be adopted by others to stay competitive. Ultimately you need associates to manage the latest tools, integrate them into workflows, etc... a managing director certainly isn't getting his hands dirty doing that. Will there be less associates hired due to AI? Maybe at the margins but I think it will be less of an impact over the next 5 years than the hype train some people are pushing.

A lot of this is overblown but I have some thoughts about Excel. You think AI will affect Excel tools?  That is one of the more unlikely outcomes considering the constraint is continued adoption of Excel in it's current format. I think AI will eventually make alternatives like Python more accessible to non-technical users and the current form of Excel will be obsoleted altogether. This is the type of disruption people should be looking at because these are the productivity gains where you can try to be an early adopter.

Jobs that revolve around networks of high-end professionals (audit/law/IB/consulting) should be worried. Your network is full of smart people who will figure out how to use AI to save costs on you. 

But if you're in one of those roles and pivot to serving smaller clients (SMBs, retail, etc.), it's a different story. Those markets are way more old-school and much slower to adapt to tech. If you actually know how to use AI to your advantage, you can dominate. These clients won’t replace you with ChatGP, on the contrary, if you leverage it properly you can get more business.

Now, assuming you think those old school guys will eventually get repalced in the long-term with milennials and new generations, this brings another opportunity. They will overrely so much on AI and will be se overconfident in it that this agians opens a door for market inefficiencies. Think a guy who sells his business with AI and prices it at 20% lower than its true value. This is an opportunity. For law, someone drafts somthing so confident that then puts him in a worse position, so he litigates. For instance, I commented on another thread about a guy having an AI-run IB. The fact that he sells businesses with that process means that I, as a PE, can screen it, see they missed key points that should make the business be worth 30%, so I can get it 30% cheaper. This to say that this intellectual laziness and outsourcing to a system, from another perspective, is an opportunity for new inefficiencies. 

As for PE, it doesn't rely on other professionals to carry their work. PE gives work to all those other adjacent fields: law/consulting/audit. The fact that those AI tools can be leveraged internally means that less work will be outsourced, and less juniors will be needed, which brings another interesting dynamic. The fact that finance people like to have high compensation. So if firms can cut 60% of their headcount, maybe at one pointin the future, in the most optimistic scenario, the industry can get back to its pre-2008 compensation packages... for PE the carry will be also better if you're saving costs on juniors and keep a leaner team of mostly seniors. Re. juniors, time to create your own opportunity and stop following an outdated handbook. Times change.

It definitely won't be modeling automation that replaces junior finance people.  Now if the AI is going to do a lot of other tasks too, then that's a conversation.  And I think that's a threat to all generic business-y white collar skills broadly, not specific at all to finance.  

So if the question is, should someone choose [insert normie smart kid job here . . consulting, law etc] over the IB/PE route because of these modeling tools, I think the answer is absolutely not.  If it's coming for finance kids, its coming for everyone else who spends most of their time in MS Office.  

The question maybe worth asking is, if it wipes out the entire concept of huge legions of smart kids making $200k+ for a good handful of years after graduation by just screwing around in MS Office, should people be looking to build more specific skills that AI can't go after.  That's a tougher question.  

Two sides of this topic are definitely overdone to me:

1. All the retards on social media posting "OMG I tried to make ChatGPT do ____ today and my mind is blown!".  Fuck those people.  They're posting for clicks and it's exaggerating the hype, especially around the speed of change.  I think long-term change will be huge but right now I can't even get the thing to give me much leverage in basic PPT.
2. The naysayers who are like "we thought the industrial revolution and internet would take jobs, and all they did was create more jobs! This will be the same!" Fuck those people too.  It's not hard to see this one is going to be different.  

It will take a generation, current (and soon to be) seniors are too stuck in their ways to replace juniors entirely 

This is the defining issue of our time right now. I have no idea why people aren't absolutely terrified of this technology. It's a net loss for society tbh, or at least under the capitalist economic model. 

If you end up losing your job because of AI, there will not be another job. That's it... that's all the money you'll make in your career...

Modeling is table stakes and not what makes a good or bad investment. Just means leaner deal teams

They just won't increase pay. PE Analysts and Associates aren't just robots, especially the good ones. Also, because AI can't replace Partners and VPs, they still need people to grow to fill those positions. Plus imagine asking LPs to invest in your fund if all of the due diligence is ran by AI. 

Just because a plane can Autoland doesn't mean you throw away the pilots. Like AI, Autopilot is a way for pilots to control the airplane without holding the yoke for 10 hours. It just allows you to do more human related tasks, like strategy and thinking.

IDK, AI is rapidly evolving and they used to say that a twin engine wouldn't be able to cross the Atlantic. I could see many firms (MM/UMM PE) reduce the size of their analyst classes, but Institutionalized AAMs (BX/KKR) would not completely close them.