Jobs and artificial intelligence — which disappear and which don't

The topic of jobs and artificial intelligence stirs up more emotion today than almost any other labour-market issue — unsurprising, given that AI tools have started writing text and code, analyzing data, and generating images within just a few years. But before you pick a degree based on "AI resistance," it's worth breaking the topic down: not every instance of automation means a job disappears, and AI often supports tasks rather than replacing them.

How job automation actually works

Labour-market researchers have long distinguished automating tasks from automating entire jobs. AI most easily takes over repetitive, rule-based activities — data entry, basic document analysis, generating standard content. It's much harder to replace tasks that require empathy, complex judgment, negotiation, or legal and ethical responsibility. That's why most professions don't disappear entirely — instead their internal structure changes: technology takes over some routine tasks, while the person doing the job focuses on tasks that require human judgment.

Jobs more exposed to change

• Simple office and administrative work based on repetitive procedures.
• Basic customer service run on rigid scripts.
• Translating standard, non-technical texts.
• Basic bookkeeping and data entry tasks.
• Some creative-industry tasks involving generating simple graphics or ad copy.

Jobs where AI is more likely to support than replace

• Medicine and psychology — AI helps with diagnostics, but decisions and the patient relationship remain a human domain (see also how much a psychologist earns and how to become one).
• Teaching and counselling — technology supports but doesn't replace the mentor-student relationship.
• Engineering and IT — developers increasingly use AI as a tool rather than a competitor; see our article on IT degree courses and how much a programmer earns.
• Jobs requiring physical presence and manual precision — trades, construction, some medical specializations.
• Management, negotiation, relationship-based sales — trust and cultural context are hard for an algorithm to replicate.

Table: indicative automation risk by professional area

Professional area	Indicative risk level	Notes
Simple administrative work	High	High task repetitiveness
Scripted customer service	High	Already partly automated
Programming and IT	Low–medium	AI as a supporting tool, changing the nature of the work
Medicine, psychology, therapy	Low	Requires responsibility and a patient relationship
Trades, manual services	Low	Physically difficult to automate

This is indicative data based on general trends described in sources such as the Occupation Barometer — the exact situation depends on the industry, company, and region. It's also worth remembering that this same data is updated every year, so the assessed automation risk for a given job can shift as technology evolves.

What this means for choosing a degree

Instead of looking for a degree that's "100% AI-proof" (because such a thing barely exists), it's worth seeking paths that combine subject knowledge with skills that are hard to automate: critical thinking, communication, project management, and basic fluency with AI tools. Our overview of jobs of the future in Poland and checking whether a degree has a future based on hard data can also help. It's also worth considering interdisciplinary degrees that combine, say, technology with the social sciences — this kind of skill combination tends to be especially resistant to automation, since it requires both technical knowledge and an understanding of human context.

How to prepare for working alongside AI, whatever your field

Whatever degree you choose, it's worth developing a few universal skills already at university that increase your resilience to technological change:

• Basic familiarity with AI tools in your field — using them thoughtfully is increasingly expected by employers.
• Critical thinking and fact-checking — the ability to judge whether an AI tool's output is correct and sensible.
• Communication and teamwork — hard to automate, and increasingly valued in roles that span multiple disciplines.
• Flexibility and a willingness to learn — the job market changes faster than it used to, so being able to quickly pick up new tools is becoming a skill in its own right.

These skills are worth developing regardless of your chosen degree — whether you study a technical, humanities, or social-science field.

How to check this for yourself

You don't have to analyze this from scratch on your own. Our free, 10-minute degree and career matching test takes into account both your aptitudes (the RIASEC model) and labour-market realities, including the indicative automation risk of specific professions.

Summary

Artificial intelligence is changing the job market, but it rarely eliminates whole professions — more often it changes how they're carried out. When choosing a degree, it's worth betting on skills that complement AI rather than trying to compete with what it does best. This is an important factor in the decision, though it doesn't replace an individual conversation with a careers adviser.