Career Implications of Agentic Coding
Obviously, we’ve seen and discussed the market’s reaction to the advent of this new technology, but I haven’t seen much discussion here about our own career trajectories yet. Thought I’d share a few thoughts.
Quick background: I work at a large LO AM in an investment-focused role with some CS background (intermediate level; rarely used daily until now). Our firm recently rolled out its own autonomous coding environment with Opus 4.6 last week…
Am I the only one having a holy shit moment?
Scripts that automatically read daily attribution and then develop a frontend. BBG IB API scraping replacing seven terminal functions in seconds. More complex signals built. NLP agents embedded directly into our datasets. Etc.
In 10 days, I’ve made tools that have eliminated around 100 hours of manual work for my pod (and this is just the low-hanging fruit stuff). These projects would have taken months of development, corporate red tape/politics, and $$$$.
I can attribute this forum to significantly helping my career, so I hope to give back and hear how others are approaching this:
- As a junior, you must at least learn how to vibe code (dare I say prompt engineering? It is an art) and understand data structures at a high level. Yes, the stickiness and UI of Excel are great, but the ceiling with programming is far higher.
-Everyone runs on a mental algorithm. Spend time getting this out of your head and into code.
-There is effectively no barrier to entry to creating software anymore. Anything software-based that you interact with, you can build with enough time and the right models. Around 75-90% of OpenAI’s Codex was developed with AI assistance. Think about that. What’s stopping you from building?
-I envision a more likely-than-not future (3 years) where BO/MO is heavily reduced and each pod has an “embedded junior” replacing many BO/MO functions via algorithms (see Block cutting ~40% of their workforce and Dorsey stating they’re early; our own firm is having similar discussions). I think FO will be fine for a while, though teams will get leaner (historically, financial services always find a way to collect a fee). Take first-mover advantage and impress your boomers.
- Am I overreacting? How have people on the buy side been using this? What do people think the societal implications will be (get up to speed on Hegseth v. Anthropic if you haven’t already)? Any tips or tricks with these tools?
-I tend to think this will make markets even more efficient and that groupthink could become even more prevalent (both from an investment research standpoint and across society).
Didn’t want this to come off as some Citrini Research doomer post, but that may have failed. All that said: be aware, be proactive, and be kind. Don’t get left in the dust in an ever changing world. Happy coding.
The career implications of agentic coding, as highlighted in your post, are profound and transformative, especially for the financial services industry. Here's a breakdown of the key points and insights based on the most helpful WSO content:
1. The Rise of Autonomous Coding and Its Impact
2. Skills for the Future
3. Shifts in Workforce Dynamics
4. Market and Societal Implications
5. Practical Tips and Strategies
6. Broader Perspective
In conclusion, you're not overreacting. The changes you're observing are part of a larger trend reshaping the financial services industry and beyond. By staying proactive, continuously learning, and embracing these tools, you can position yourself for success in an ever-evolving landscape. Happy coding indeed!
Sources: Will robots replace your consulting or financial career?, New breed of Finance and Capital Markets Professionals, Learning coding on your own for finance?, Q&A: Managing Director at Large Global Asset Manager, Work at top AM -- AI will virtually kill of entry-level roles in AM over the next 5yrs
I disagree with your assertion that BO/MO will be significantly harmed in the short term. I expect them to resist the AI wave for quite a bit longer.
I would share my opinion on the Front Office as well, but I am not able to disclose since I am working on AI related projects for them right now
Appreciate the response here and the varying opinion.
At least personally, I see this in two stages:
1) building the infrastructure (cleaning and organizing the data, getting more APIs, creating agents and sub agents, corporate red tape, etc)
2) Integration into investment decisions
We’re at stage 1 still (and sounds like you/others I’ve spoken with are as well?) are at. What’s your timeline on this buildout industry wide or do you believe this is where BO/MO perhaps could come in? How strong are the moats that you described in points 2 and 3 going forward? And where do juniors fit in?
What are some of task you’ve automated?? 100 hours of work gone is impressive. I’ve been going ham at it all week building dashboards etc but feel like it’s just made things easier to visualise / navigate, but ultimately you need the underlying data.
Aliquam et aspernatur odio. Saepe laboriosam molestiae non aperiam. Aliquam sunt sit tempore earum voluptatum placeat.
Molestiae vel commodi sit ducimus nam. Harum aut vitae nihil. Inventore ipsam consectetur eos odit cum veritatis atque cumque. Est itaque cum illum dignissimos voluptatem eveniet dignissimos. Et et unde commodi beatae temporibus ratione. Quisquam aut quia consequatur corrupti sed voluptatem esse quod. Voluptatem occaecati similique labore totam.
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...