Will artificial intelligence make me unemployed?
I am a creative spirit – I have been working with and on the web for more than 25 years: My daily work includes things like digital communication concepts, design, and content strategy, as well as front-end and back-end development, CMS integration, and accessibility. I enjoy this diversity – and I connect people, ideas, and technology with a great deal of empathy.
In recent months, I have been experimenting with all kinds of AI tools – image generators, code generators, and pure language models. All of these tools have an impact on my work. And at some point, I was gripped by fear: Could AI replace me?
When it comes to programming, I was initially excited about the possibilities; my prompts resulted in functioning code snippets. I no longer have to worry about complex regular expressions either. (It doesn't matter what that is, but it's definitely something that makes some brains smoke.)
Soon I was certain: it wouldn't be long before I would be superfluous.
Falling asleep became increasingly difficult, and the nights grew shorter. I made many of my colleagues nervous with long phone calls, and I could already see myself unemployed—not a particularly pleasant prospect.
The more I looked into AI, the more I realized that things are not as bad as they seem.
Language is the means of communication for almost all humans. Whether written or spoken, this is precisely why AI is based on language models. Language models are mathematical systems that interpret, analyze, and generate language. They do not understand language, but rather recognize patterns and operate with probabilities. Familiar content is recombined, creating something that appears to be new.
The internet is currently being flooded with all this seemingly new content. On LinkedIn, you come across posts created with AI much more frequently. AI-generated posts are easy to spot. In most cases, you don't even need to read the text; the icons used immediately reveal the origin.
Artificial Intelligence also supports me in my Work.
At school, I missed out on learning how to use commas, but today, thanks to AI, this deficit is no longer an issue. I fight for my convictions, but sometimes I overlook certain aspects in my arguments, and AI points them out to me. It helps me find suitable headlines, provides code snippets, and assists me in creating a basic framework for a CMS extension. It's all incredibly practical and saves time.
Programming is a very technical field. AI can recognize patterns and structures and generate new code by recombining known code modules. However, this is also problematic because AI generates code based on statistics, not logic. It tends to reproduce known patterns and apply them in other places as well.
Example 1: AI also reproduces incorrect information
When I ask ChatGPT about accessible navigation, I often receive the following code snippet:
<nav id="mainNav" role="navigation" aria-label=“Main navigation”>
However, the HTML element <nav> already implicitly has the ARIA role “navigation.” An additional role="navigation" attribute is therefore superfluous. Assistive technologies automatically recognize that this is a navigation area – even without the explicit role attribute. It is not surprising that such errors or similar ones are frequently found on the internet today. The technical fundamentals are lacking, and incorrect information is simply copied.
Example 2: AI is forgetful
Recently, I used AI in my development environment to program a Joomla! (CMS) extension. I was almost done and wanted to make a small change when suddenly nothing worked anymore. What a bummer.
Just because the AI had decided, shortly before reaching its goal, to partially change the names of my variables in the middle of the process. This change was a side effect of the “guessing process”: it simply forgot (did not save) the original name of the variable.
Finding the error here requires human expertise and a deep understanding of the subject matter, so I'm not superfluous after all.
Example 3: Depending on the situation, it can be costly and time-consuming.
I had envisioned it so beautifully: completing routine tasks quickly while I took care of something important.
The goal was to replace outdated functions in my Joomla templates. This included, in particular, removing the “J” prefix from class names and using use statements to correctly integrate namespaces.
I set the AI to work its way through around 24 files. However, it only managed one file, wrote the corresponding tests, and then reported: “Your credits are all used up.” Since I didn't know exactly how many I had already used, I quickly bought 10 new credits for 20 euros so I could continue.
I started again and watched the clock: my AI took about 3.5 minutes per file. I let it continue, but after a few files, the message “Credits are all gone” appeared again. That got me hooked: I had already spent 20 euros and was competing against the AI. I needed about 1 to 1.5 minutes per file with “Find and Replace,”.
So I was significantly faster and, above all, significantly cheaper.
Analyzing individual files is quite costly for AI. In most cases, billing is not per file, but per request or per number of characters. That's why I now check the consumption per step to avoid unpleasant surprises.
At the moment, the revenue of some AI companies is still modest. As users, we feed our data into the free tools, and as soon as we become truly dependent on them, prices will probably rise significantly. People or companies with lower incomes will be left behind, as has always been the case.
Using AI is an educational task.
But what really worries me is the danger that people will lose skills in their respective fields or fail to develop them in the first place due to uncritical use of AI. Those who rely too heavily on AI run the risk of acquiring only superficial knowledge and losing the ability to understand complex relationships for themselves.
Our brain functions like a muscle; if it is not exercised, its performance declines. Concentration and memory deteriorate.
The study “Your Brain on ChatGPT” from June 2025 found significantly reduced neural connectivity in the brain (EEG) and reduced memory in ChatGPT users. (Study: Your Brain on ChatGPT, MIT 2025)
In the future, this could have social, economic, and health implications, the consequences of which we cannot yet foresee.
The current hype surrounding artificial intelligence is also noticeable in German companies. Many companies have integrated various types of AI into their everyday work, if only to avoid being considered backward or resistant to innovation.
It is often overlooked that the mere use of AI is not an end in itself; it requires a strategy and trained users. AI can automate routine tasks, accelerate processes, and open up new business opportunities.
As long as you do not operate your own AI, I recommend caution: most tools process user input to generate responses and store this data to improve the model. Personal data, passwords, or confidential business content should never be entered, as they could theoretically be used for analysis or training.
In the European Union, the use of such services is subject to the strict requirements of the GDPR, which regulates how personal data may be processed. Although providers such as OpenAI use encryption and security protocols to protect the data, the data is still uploaded to their servers. Therefore, both companies and individuals should carefully consider what information they entrust to AI and, in a corporate context, develop internal guidelines for the secure and privacy-compliant use of AI tools. When I observe my private environment, I see that there is still a great deal of need for education—and it is likely no different in companies.
The key point is that we should all see AI as a tool. It can serve to expand our knowledge, but not replace it. Those who actively question, analyze mistakes, and challenge themselves not only maintain their motivation, but also train themself. AI is a complex topic that affects society, the economy, politics, and the environment in equal measure. We should use it to our advantage without becoming dependent on large corporations, while taking care not to burden our environment too much with high energy consumption.