Is artificial intelligence taking jobs, or does it offer a new system and approach to work tasks? The impact of AI on jobs is discussed in an interview for aktualne.cz by Alexandr Vasilenko, a certified Microsoft instructor at OKškolení.
AI represents a revolutionary tool with the potential to transform work across many fields. It is no longer a complex technology to be used by specialist individuals, but a practical skill that is becoming essential for a growing number of people. Do you want to stay competitive in the job market or increase your work efficiency? Then you should learn the basics of AI for everyday use.
Who needs to know how to work with AI today?
Very generally speaking, anyone who is capable of critically evaluating its output. AI is commonly used by marketers to generate content, by sales professionals to gather information about contacts, or by finance specialists to process data in Excel. Its practical applications span industries and hobbies alike, from tips for a healthy lifestyle to complex economic analyses. However, it is important to keep in mind that AI often guesses answers or even makes them up – and creates so-called hallucinations.
So it’s not just a tool for IT experts?
Definitely not. Tools like ChatGPT or Copilot are designed for a wide range of users. Even so, people must act as a control mechanism. Artificial intelligence will almost never answer “I don’t know”; it is not programmed that way. That is why we should never be satisfied with the first output. It is up to the user to critically assess the correctness of the answer and ask follow-up questions. In my trainings, I therefore always teach participants to ask questions such as “Play the role of a critic and evaluate your own answer” or “Now you are an economics expert – suggest improvements to the answer.” This approach makes it very easy to achieve higher-quality results.
Which specific tasks can we already delegate to AI today?
AI can significantly speed up routine tasks at the level of junior positions. Among the most practical tasks that AI handles effectively are creating draft emails and presentations, summarizing meeting notes, analyzing data, quickly extracting key points from large articles, and of course translations. Those who know how to ask better questions – in other words, how to prompt – have an advantage. Most often, AI is used for text summarization or content generation, whether text or code. Advanced users also make use of AI agents that can be trained for specific company tasks.
Can you give concrete examples of such agents and what they are used for?
You can think of an agent as a specialized AI function that can be created, for example, in more advanced versions of ChatGPT or Copilot. The main benefit lies in automating repetitive and routine tasks. For instance, you can set up an agent whose task is to create a summary of economic news and stock market updates from a specific portal weekly, let´s say Monday mornings, fill relevant data into a predefined table, and automatically send it to the entire sales team. Another agent may focus on publishing social media posts according to an approved plan or creating draft PR articles for a new campaign. Crucially, the quality of their work is directly proportional to the level of detail in the prompt, so it pays to invest real effort in defining the task. This can save companies time on routine and administrative activities.
Where does AI reach its limits and fail to meet expectations?
The most common dissatisfaction stems from poorly formulated questions. Tools like ChatGPT or Copilot are based on large language models that, using vast training datasets, propose only a probably correct answer. If the question is too general – for example, “How can I best prepare for a job interview?” – the answer will be vague. The quality and relevance of the output improve dramatically only when we provide AI with maximum context, such as the job title, the company, previous experience, or even the wording of the job advertisement itself.
Another limitation is energy consumption and processing costs. This is why free tools have limited functionality, and generating demanding tasks such as images or videos takes longer, with outputs often containing errors and imperfections. The reason is that AI models would require significantly more time and computing power to achieve near-perfect results, which is not commonly available for economic reasons.
Where and why do companies fail in implementing AI?
A very common reason is unrealistic expectations. I ask a simple question, get an insufficient answer, and don’t feel like experimenting further to see how it could work better. Another reason may be employees’ natural resistance to change, which takes time and patience to overcome, along with insufficient ongoing training. Companies often focus only on safe usage, not on the real possibilities and opportunities AI offers for improving efficiency. AI is certainly not a cure-all; it will not solve structural problems within a company or fix an uncompetitive product.
Where should I start if I have zero previous experience with AI?
There are many courses to choose from. For complete beginners, I recommend the Copilot in Microsoft 365 course, which introduces the benefits of the tool through practical examples in the course of one morning. For those with basic awareness, the AI Work-life Mastery course is suitable, where we focus in more detail on ChatGPT and Copilot at an intermediate level. Advanced users can then focus on automation tools such as Microsoft Power Automate, where they learn to set up recurring tasks without programming, such as automatic email sorting or report generation. Beyond courses, self-study is key – constant experimentation, practice, and a critical approach to outputs are the only way to achieve reliable results.
So a course is not the end – practical follow-up training is also important…
Continuous education is crucial because artificial intelligence is now connected to almost every field. As a regular user, you cannot be satisfied with just basic training; you need to learn responsible usage and advanced features. These include the aforementioned AI agents, which you can quickly train for a specific project or task, pushing work efficiency beyond basic usage. Ideally, AI implementation in a company should also be supported by AI ambassadors who inspire and spread knowledge across the organization. At OKsystem, for example, this role is fulfilled by OKlab – a research and technology hub where we experiment with AI use on our own on-premise hardware.
What should be said in conclusion about AI?
It is time to let go of the fear that artificial intelligence will take your job. Instead of worrying about threats, focus on the enormous efficiency, time savings, and competitive advantage that AI can bring. The rule is clear: you don’t need to fear that AI will replace you, but you should worry about a colleague who has learned how to work with it. Don’t wait until this essential skill catches up with you. Investing in quality education in AI is an important path to staying relevant and successful.