{"id":14319,"date":"2026-05-19T14:40:02","date_gmt":"2026-05-19T12:40:02","guid":{"rendered":"https:\/\/manualjobsearch.com\/?p=14319"},"modified":"2026-05-19T14:45:51","modified_gmt":"2026-05-19T12:45:51","slug":"wie-ki-die-datenberufe-spaltet","status":"publish","type":"post","link":"https:\/\/manualjobsearch.com\/en\/wie-ki-die-datenberufe-spaltet\/","title":{"rendered":"How AI is splitting data professions: Who wins \u2014 and who disappears?"},"content":{"rendered":"
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\n
\n\nAn experienced Director of a 600-strong data organisation speaks openly about the new reality: Most junior data analysts have hardly any chances. Mid-level professionals are under pressure. But whoever thinks correctly is experiencing the best year of their career. What you can do specifically.\n<\/i>\n\n

The uncomfortable question: Are you a \"human-machine interface\"?<\/h2>\n\nThere's a question that's been shifting the mood in many data teams over the past few months. It's as simple as it is harsh:\n\nIf the business asks for a metric \u2013 you take the request, write an SQL statement, and return the number. What intellectual contribution have you made?\n\nThe answer is: none. You were a translator between a business question and a programming language.\n\nA seasoned data director puts it bluntly:\n\n
\"If that's all you do, then I'm sorry. You should seriously think about your future.\"<\/div>\n\nThis is not cynical. It's the sober observation of someone who is currently running an experiment in their organisation: AI agents versus human analysts. And the result will surprise you.\n\n

The test that changed everything<\/h2>\n\nA specific business task: pricing for warehouses. The particularity of this task has a catch \u2013 if a warehouse is 50 percent occupied, every additional sale is practically pure profit. The warehouse is already there anyway. However, if the warehouse is 100 percent full, prices must be increased.\n\nIn other words: the most important thing you need to find out first is the utilisation rate. Everything depends on it.\n\nWhat happened? A team of eight analysts worked on the problem for a week. They suggested clever actions: a promotion here, a discount there, a bundle somewhere else.\n\nThe Director asked: \"Great, people. But what's the utilisation rate?\"\n\nAnswer: \"We haven't calculated that.\"\n\nIn parallel, AI agents worked on the same task. And the very first thing they did was calculate the utilisation rate. Because, logically, that's the most important question.\n\n
Final score: AI agents 1 \u2014 Humans 0.<\/div>\n\nThat doesn't mean the agents are smarter. Their analysis quality was mediocre. But they asked the right question - and the team missed it.\n\n

What this story really means<\/h2>\n\nAI divides all data professionals into two camps today \u2013 and the line is not where you might expect.\n\nIt's not about whether you can use ChatGPT. It's about whether you contribute intellectual added value.\n\nModule 1: Human-Machine Interfaces.<\/strong> These people take business requests and translate them into code. The entire thinking has already been done by the business beforehand \u2013 the key figure is known, the use case is clear. It just needs to be executed.\n\nThis role is at high risk. AI is already doing this today, faster and without a break.\n\nUnit 2: Thinker.<\/strong> These people first ask: Why do we need this number at all? What happens to the result? What are the scenarios? What question are we currently overlooking? This role becomes more valuable \u2013 not less.\n\n
You have to ask yourself an honest question: is my work intellectual or quasi-intellectual? If the real intellectual component was low \u2014 then things are not looking good.<\/div>\n\n

A simple self-analysis: What do you actually do all day?<\/h2>\n\nThe practical tool is as simple as it is uncomfortable:\n\nSit down for one to two weeks and write down what you do daily. Do you talk to the business? Do you write code? Do you think? Do you design systems? Do you have meetings?\n\nThen, for each item, ask yourself: Can't an AI do this quite well already?\n\nFind your bottleneck. What's preventing you from becoming faster?\n\nUp until now, the bottleneck for many has been writing code. Today, this can be solved with AI \u2013 you stop getting stuck there.\n\nBut then comes the serious question: What's left? If the code is gone, what do you do? If you don't have an honest answer to that, you now know what you need to work on.\n\n

The situation by career level: unvarnished<\/h2>\n\n

New starters: honest words<\/h3>\n\n
I'm glad I don't have children aged 18 to 20 who have to start their careers now. I have no answer for them.<\/div>\n\nBig data organisations today are practically not hiring junior staff anymore. If interns are taken on, it's only those who have the level that would have previously been expected of mid-level employees.\n\nAbout one in a hundred interns meets this new standard. This one person is inundated with offers. The other ninety-nine cannot find employment.\n\nWhat distinguishes this one person? They can design a system. They have realised their own projects. They have built something that works.\n\nWhat stops anyone from just building a web service, gaining a few hundred users and working with it? Nothing. But out of a thousand people who understand this, ten to twenty do it.\n\nIt's about something called initiative. Ironically, in the AI agent era, companies are looking for people with initiative. With drive. With the ability to tackle something themselves without someone saying, \"Do this.\"\n\nThis is not a learnable skill \u2013 it's a character trait. And the result is a harsh divide: a small group is overwhelmed with offers, while the large majority finds nothing.\n\n

Mid-Level: Probably hit the hardest<\/h3>\n\n
\u201eThese people thought they'd made it. They're in. They're established. And now the heavy boot of AI is crushing their fingers.\u201c<\/div>\n\nMid-level professionals face a particularly unpleasant task: honestly analysing their own work. Where are they really spending their time? Which of that is valuable?\n\nHere's a little-known phenomenon from large tech corporations: many engineers don't understand what's truly important. But they know they're being evaluated. So what do they do? The most rational strategy: do as much as possible. The lottery principle \u2013 produce a lot in the hope that something valuable emerges.\n\nThis leads to burnout. And it doesn't lead to better results.\n\nThe way out: honest self-analysis. Why did I do this? Was it valuable? What really matters? Who is holding me back?\n\nThis is tough. Not everyone can manage it. Those who do are rewarded. A mentor or an external consultant can help here \u2013 it's easier to see problems in another person than in oneself.\n\n

Senior-Level: Life is getting better<\/h3>\n\nReal seniors \u2014 those who have always been autonomous units, responsible for a system or process \u2014 have it easier today, not harder.\n\nPreviously, seniors delegated tasks to junior and mid-level employees \u2014 broken down into small pieces, almost like to an agent. Today, these tasks can go to AI agents. The senior's life has become easier.\n\nWhat was previously impossible \u2014 for example, being able to test one-twentieth of all hypotheses \u2014 can now be done with one-fifth. This is a significant leap.\n\nBut be warned: it is a myth to believe that a senior will achieve three times as much as a result. A senior's main bottleneck is not their bandwidth \u2014 but rather internal bureaucracy and the absurdity of large corporate processes. AI changes nothing about that.\n\n

Leaders: A new, uncomfortable question<\/h3>\n\nThe task of managers has not changed \u2013 they are responsible for results. But a new question arises:\n\nIf I can get done everything that is expected of me in 15 hours a week \u2013 why should I work the remaining 25 hours for free?\n\n
\"I work 15 hours a week. The other 25 hours are essentially for free. Because everything that was expected of me was finished in the first 15 hours. Some people say: No thank you, that's enough for me.\"<\/div>\n\nThis inevitably leads to the conclusion that performance appraisal systems need rethinking. The old model \u2013 \"everyone gets more or less the same\" \u2013 no longer works.\n\nTop performers must be visibly rewarded. Major companies are already reacting: bonuses up to 300 percent of salary for the best employees. Significantly increased share packages.\n\nAnd middle management is being drastically reduced. If every senior is a highly productive autonomous unit with agents, why do you need so many managers? The first companies are introducing \"Giga-managers\": instead of 5 direct reports, now 10 or 15. No more weekly one-to-one meetings.\n\n

What is being asked for \u2013 and what is becoming a commodity<\/h2>\n\nHere's a clear separation:\n\n
    \n
  • Commodity (everyone can do it)<\/strong> Writing code. Still necessary, but no longer distinguishing.<\/li>\n
  • Critical (few can do this):<\/strong> Design systems. Understand trade-offs. Keep a whole architecture in mind. Formulate thoughts clearly.<\/li>\n<\/ul>\n\nThe question \"which technologies should I learn?\" is less important today than it seems:\n\n
    Technologies aren't crucial. Today one, tomorrow another. You must be able to solve problems. Build some web service. Get 100 users for it. Work with it. You'll learn the necessary technologies on the side.<\/div>\n\nBut one thing is mandatory: every web frontend for an LLM (chat interface) \u2013 that's standard. And an agent-assisted development environment (various tools are available today) \u2013 you must master that now. That is not optional.\n\n

    Where is the real money being made right now \u2013 classic ML or GenAI?<\/h2>\n\nAn important insight that often gets lost in public debate:\n\n
    The vast majority of money is still being made with classical machine learning. The whole AI agent economy is based on one idea: we're increasing productivity. What is productivity increase? It's the ability to do the same thing with fewer people. The whole GenAI economy is based on us firing you all.<\/div>\n\nConcrete success story: A large European fintech company has switched three-quarters of its customer support to AI. And the support is actually working at a good level \u2013 most customers no longer need a human agent.\n\nBut that's customer support \u2013 not the all-changing revolution the market expects.\n\nPrediction: Agents will replace ten million jobs in the back offices of large companies within 3 to 5 years \u2013 especially in India, Malaysia and similar hubs. But the return on investment for AI integration in companies is practically zero today, and won't be there next year either.\n\n

    How to prepare - three steps to your next job<\/h2>\n\nIn large data organisations, a typical application process for senior positions looks like this today:\n\n
      \n
    1. Code Interview.<\/strong> Two formats. Either you get code for review: \"What's good, what's bad, what would you change?\" Or a code snippet with the task of implementing an ML algorithm. Google and even LLMs are permitted for support \u2013 but not for complete writing. One observes how the candidate approaches problems.<\/li>\n
    2. System Design.<\/strong> Also two formats. \"Greenfield\" \u2014 design a system from scratch. \"Brownfield\" \u2014 talk about a system you've built and how you would change it today.<\/li>\n
    3. Behavioural<\/strong> How do you behave in certain situations? How do you talk about your experience?<\/li>\n<\/ol>\n\nThe most common mistake candidates make: not thinking their final answer through, but changing it as soon as someone asks, \"Is that really your final answer?\" The problem isn't a wrong answer. The problem is that the candidate hasn't done any critical self-assessment.\n\n

      How you stand out today<\/h2>\n\nThere is a sobering realisation: to belong to the top five percent, no genius is needed.\n\nThe instruction is mundane: Analyse. Observe how successful people do it. If necessary, ask a consultant or mentor. Adapt. Apply. Repeat.\n\n
      Remarkably, many people consciously do everything to avoid self-improvement. Even after this explanation, 5 out of 100 will implement it. And if it were 20, the competition would be four times as hard.<\/div>\n\nSpecifically, this means that you can stand out in the market by:\n\n
        \n
      • Top-tier data science competitions<\/li>\n
      • Olympics<\/li>\n
      • Meaningful open-source contributions<\/li>\n
      • Publications, lectures<\/li>\n
      • Own projects with real users<\/li>\n<\/ul>\n\nThe content is not crucial. The proof is: this person latches onto something and sees it through to completion.\n\n

        Would you join the profession today?<\/h2>\n\nThis question, put to an experienced Data Director, yields a remarkably honest answer:\n\n
        I'm not sure I'd get into data science today, to be honest. It's a red ocean. It's death.<\/div>\n\nWhat was different back then? Back then it was easy to get into, a lot of money, significantly better than the work he had done before. Today the conditions are different.\n\nAnd where would he go instead? An honest answer: he doesn't know. \"The classic problem of a person at the beginning of their career.\"\n\n

        Which data field is the most secure?<\/h2>\n\n
        Scenario analysis and risk assumption. Risk assumption does not absolve AI.<\/div>\n\nRisk is responsibility. And responsibility must be borne by a human being \u2013 not a machine.\n\nWho decides to grant a loan? To pay out on an insurance policy? To release a medication? To make an investment? Who bears the legal, ethical, and personal consequences if the decision was wrong?\n\nThese questions are not answered by machines. They will not be answered by machines.\n\n

        Conclusion: It's about new thinking, not new tools<\/h2>\n\nThe key takeaways from this market observation:\n\n
          \n
        • AI does not replace the data professional. It replaces the translator between business and code. Anyone who does more than that is safer than ever before.<\/li>\n
        • Tools are secondary. Which library, which framework \u2013 it hardly matters. What counts: identifying problems, designing systems, understanding trade-offs, asking the right question.<\/li>\n
        • The most important step today: honest self-analysis. What am I really doing? Which of that has intellectual added value? What is my bottleneck?<\/li>\n
        • And build something. Not talk, not plan \u2014 build. A web service. A tiny tool. A data analysis with real consequences.<\/li>\n<\/ul>\n\nProve \u2013 to yourself first \u2013 that you can finish something.\n\nIn a market that is changing so rapidly, the most valuable signal is not a certificate. It is the proof that you can think for yourself.\n\nEditorial note:<\/strong>\nThis article is based on an expert interview with a seasoned executive in Data and AI at an international corporation. The assessments and observations presented reflect the personal position of the expert.\n<\/article>\n\n\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>","protected":false},"excerpt":{"rendered":"

          Ein erfahrener Director einer 600-k\u00f6pfigen Daten-Organisation spricht offen \u00fcber die neue Realit\u00e4t: Die meisten Junior-Datenanalysten haben kaum noch Chancen. Mittlere Fachkr\u00e4fte stehen unter Druck. Aber wer richtig denkt, erlebt das beste Jahr seiner Karriere. Was Sie konkret tun k\u00f6nnen. Die unbequeme Frage: Sind Sie ein \u201eMensch-Maschine-Interface\u201c? Es gibt eine Frage, die in den letzten Monaten […]<\/p>","protected":false},"author":1,"featured_media":14320,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[],"class_list":["post-14319","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ohne-rubrik"],"acf":[],"_links":{"self":[{"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/posts\/14319","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/comments?post=14319"}],"version-history":[{"count":7,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/posts\/14319\/revisions"}],"predecessor-version":[{"id":14329,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/posts\/14319\/revisions\/14329"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/media\/14320"}],"wp:attachment":[{"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/media?parent=14319"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/categories?post=14319"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/manualjobsearch.com\/en\/wp-json\/wp\/v2\/tags?post=14319"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}