Image by author
, Introduction
Are we all in a race to reach our self-created bottom line? Data professionals have been employed for years to develop large language models (LLMs).
Now, the number of open data positions is decreasing daily. Most of what is advertised seems quite disappointing.
By extremely low I don’t mean very low salaries or unreasonable technical expectations from candidates. No, I mean those vague phrases: “comfortable working with AI productivity tools,” “able to ship high volumes of code,” or “strong prompt-engineering skills a plus.” Translation: A chatbot is your main coding partner, there will be no advice, no standards, just brainstorming code.
A chatbot, our own creation, is now reducing us to copy-pasters of its output. This doesn’t seem like very meaningful or fulfilling work.
In this environment, is it still possible to find meaningful work?
, What is vibe coding?
Andrej KarpathyOne OpenAI Co-founder coined the term “vibe coding”. This means you don’t code at all.
What you do: You’re sipping your matcha latte, vibing, giving orders to a coding chatbot, and copy-pasting its code into your code editor.
What a chatbot does: It codes, checks for errors, and debugs the code.
What you don’t do: You don’t code, you don’t check for errors, and you don’t debug the code.
How does such work feel? Like full-time brain rot.
what did you expect? You handed over all the interesting, creative and problem-solving aspects of your work to a chatbot.
, Vibe coding has devalued coding
Andrzej Karpathy said of Vibe Coding, “It’s not too bad for tedious weekend projects, but still quite entertaining.”
Despite this, companies you’d trust – ones that don’t think of their products as “throwaway weekend projects” – decided it’s still a good idea to start practicing vibe coding.
AI coding tools came in and data professionals were driven out. For those that are left, their main function is to interact with chatbots.
Work gets done faster than ever before. You meet deadlines that were previously impossible. Your ability to pretend to be productive has taken on a whole new level.
outcome? Half-finished prototype. Code that breaks in production. Data professionals who don’t know why the code isn’t working. Hey, they don’t even know why the code is there Is working.
Prediction: Professionals who really know how to code will soon be back in fashion. After all, someone has to rewrite the code the chatbot has written “so fast.” Talk about efficiency. Well, you can’t get more efficient than that.
But how will you survive until then?
, How can you find a meaningful job now?
The principle is very simple: do the work that a chatbot can’t do. Here’s a comparison between what AI can’t do and what you can do.
Of course, doing all this requires some skill.
, essential skills
These skills are needed to find meaningful work in the age of vibe coding.
, 1. Technical Specification Writing
Most of the requests you deal with come with incomplete and unclear information. If you can turn that information into an accurate technical specification, you will find it valuable to prevent conflicting assumptions and expectations from development work. Technical specifications help align all teams participating in the project.
Here’s what this skill involves.
resources:
, 2. Understanding Data Flow
Systems don’t fail just because of bad codes. Arguably, they fail more often because of misconceptions about the data.
No matter the vibe coding, one still needs to understand how the data is generated, modified, and consumed.
resources:
, 3. Production debugging
Cannot debug in LLM production. That’s where you come in, with your knowledge of interpreting logs and metrics to diagnose the root causes of production incidents.
resources:
, 4. Architectural logic
Without understanding their architecture, systems will be designed to work in production (fingers crossed!), but under real traffic they will often fail.
The architectural logic determines the reliability, latency, throughput, and operational complexity of a system.
resources:
, 5. Schema and contract design
Poorly designed schema and definitions of how systems communicate can cause a domino effect: cascading failures that cause excessive migration, resulting in coordination friction between teams.
Create a good design, and you’ve created stability and prevented outages.
resources:
, 6. Operational Awareness
Systems always behave differently in a production environment than in development.
As the whole idea is to make the system work, you have to understand how components degrade, how failures occur, and what and where the bottlenecks are. With that knowledge, the transition between development and production will be less painful.
resources:
, 7. Necessity negotiation
“Prevention is better than cure” applies here too. If the requirements were poorly defined initially, you can expect almost endless outages and rewrites. It is difficult to attempt to improve a system once it is in production.
To prevent this, you have to skillfully intervene in the early development stages to adjust the scope, communicate technical constraints, and transform vague requirements into technically feasible ones.
resources:
, 8. Review of code of conduct
You need to be able to read code not only for its functionality but more broadly for its system impact.
This way, you’ll be able to identify risks that don’t show up in linting or tests, especially in AI-generated patches, and prevent subtle bugs that would otherwise mess up your production.
resources:
, 9. Cost and Performance Decisions
Your work has financial and operational implications. You’ll be more valuable if you consider computer usage, latency, throughput, and infrastructure bills in your work by considering them and understanding them.
This is given far more importance by companies than creating expensive systems that don’t work.
resources:
, Real Jobs That Still Feel Worthwhile
Finally, let’s talk about real jobs that still involve using at least some or all of the skills we discussed earlier. The focus may be shifting away from coding, but some aspects of those jobs may still seem worthwhile.
, 1. Data Scientist (the real kind, not just for notebooks)
AI can generate code, but data scientists provide structure, logic, and domain understanding to vague and, often, ill-formulated problems.
, 2. Machine Learning Engineer
AI can train a model, but what about data preparation, training pipeline, serving infrastructure, monitoring, failure handling, etc.? That’s the job of a machine learning engineer.
, 3. Analytics Engineer
AI can write SQL queries, but analytics engineers are the ones who guarantee correctness and long-term sustainability.
, 4. Data Engineer
Data engineers are in charge of the reliability and availability of data. AI can transform data, but it cannot manage system behavior, upstream changes, or long-term data reliability.
, 5. Machine Learning Ops/Data Ops Engineer
These roles ensure that pipelines run reliably and models remain accurate.
You can use AI to suggest solutions, but performance, system interactions, and production failures still require human oversight.
, 6. Research Scientist (Applied Machine Learning/Artificial Intelligence)
AI can’t really bring anything new, especially not new modeling approaches and algorithms; It can simply repeat what already exists.
Anything else requires specialist knowledge.
, 7. Data Product Manager
This job description is to define what data or machine learning products should do, which includes translating business needs into clear technical requirements and aligning the priorities of different stakeholders.
You can’t use AI to negotiate scope or evaluate risk.
, 8. Governance, compliance and data quality roles
AI cannot ensure that data practices meet legal, ethical and reliability standards. Someone needs to define the rules and enforce them, for which governance, compliance and data quality have roles.
, 9. Data Visualization/Decision Science Roles
For any purpose, data needs to be linked to decisions. AI can generate charts as it wants, but it doesn’t know what matters to make decisions.
, 10. Senior Data Roles (Principal, Staff, Lead)
AI is a great helper, but it’s a terrible leader. More precisely, it cannot lead.
to decide? Cross-domain leadership? Guiding technical direction? Only humans can do them.
, conclusion
Finding meaningful work in the age of vibe coding is not easy. However, coding isn’t the only thing data professionals do. Try searching for job ads, even if they require vibe coding, as well as some skills that AI still can’t replace.
Nate Rosidi Is a data scientist and is into product strategy. He is also an adjunct professor teaching analytics, and is the founder of StratScratch, a platform that helps data scientists prepare for their interviews with real interview questions from top companies. Nate writes on the latest trends in the career market, gives interview advice, shares data science projects, and covers everything SQL.
