BERLIN: When was the last time you wrote an important email without using artificial intelligence (AI)?
Did you actually read through that 46-page PDF for your project or get a chatbot to sum up the key points?
Surveys show 44% to 52% of UK workers use AI at work, though their regular or daily use remains relatively low (about 15% to 35% of those with access).
Most workers who use AI are mainly experimenting with it, or using it for routine administrative tasks.
In Germany, some 80% use AI at work at least occasionally, a study found. Some use it for important content but mainly for ease.
Many turn to ChatGPT, Gemini, Claude or other chatbots simply for convenience, says the survey by the Pinktum Institute.
It found 43% use AI at work partly because they "rarely feel like delving deeper into topics.
Are we increasingly leaning on AI to do our thinking for us at work?
Does that make us less and less competent as a result? We spoke to a specialist about deskilling caused by AI.
What exactly does ‘deskilling’ mean?
Deskilling describes the potential loss of skills resulting from the increasing use of AI, says Josephine Hofmann, head of the Collaboration and Leadership team at the Fraunhofer Institute for Industrial Engineering IAO in Germany.
There are two kinds, she says. People may lose skills they already possess. But she says the more problematic issue is that certain skills are no longer developed at all due to the use of AI systems, such as reading, understanding, and critically evaluating texts, or writing them by yourself and so generating new knowledge.
How significant is this effect?
"The findings are mixed; the research landscape is extremely dynamic,” says Hofmann. A causal link between AI use and reduced critical thinking skills has not yet been conclusively demonstrated.
But research does show clear correlations. A study by the SBS Swiss Business School identified a strong negative correlation between frequent AI use and critical thinking skills.
It is precisely fundamental skills that can weaken in the long term, such as understanding texts, evaluating information, formulating your own thoughts, and so acquiring and producing new knowledge. "This is an activity that has to be learnt and practised,” she says.
If AI increasingly takes on such tasks, it is this practice that is lacking. She says we also lose the sense of having worked something out for ourselves, which can affect our motivation and willingness to learn.
Does our own convenience play a role in this?
Yes, AI saves time and effort, which makes it so attractive. The rapidly developing possibilities can tempt us to think less for ourselves. "People are gradually losing cognitive engagement,” says Hofmann. However, one reason for this is also the "incredible pressure to speed up” in the world of work.
What risks does this create?
In the Pinktum survey, a good 70% of respondents state that they "usually" check AI results. Simply relying on the results "would be disastrous. That carries the risk that we will gradually lose key skills," says education scientist and HR expert Carina Ebli-Korbel.
Hofmann says people "tend to move from understanding towards delegating”. A key risk when tasks are outsourced to AI is that, as skills erode, AI-generated errors or inconsistencies become harder to spot. "Then you’re not really in a position to carry out quality assurance,” she warns. Blind spots can emerge, for example in arguments or analyses.
Can we counteract this?
Yes. The key lies in the conscious use of AI. "We can now certainly use AI to support our own thought processes,” says Hofmann. She recommends viewing AI as a thinking partner. Try prompting, "Challenge me, ask me questions again,” she says, describing one method.
At the same time, it is important to practise basic skills regularly without AI, even if this takes more time. Employers must provide the necessary time and managerial resources to make that possible.
Ebli-Korbel says the responsibility lies primarily with the individual. Organisations can only raise awareness. She advises professionals to consciously avoid outsourcing certain tasks to AI. "I can, for example, make a point of drafting emails myself on a regular basis. Especially when it’s about thinking and not just about phrasing.”
At work, we should also consciously consider the objective with which we approach a task. "If I want to consolidate knowledge for myself, I have to read for myself and reflect on what I have learnt,” says Ebli-Korbel.
Can AI also strengthen our skills?
"What you can lose, you can also gain,” says Ebli-Korbel. For example, creativity can be fostered by taking AI-generated results as a starting point and developing them further independently.
Targeted use of AI can help to expand cognitive abilities, says occupational psychologist Markus Langer in an interview with German psychology publication Psychologie Heute, partly because increased attention is required to spot errors in AI-generated texts. Those who critically scrutinise and supplement AI suggestions are training their own judgement.
AI can also help people get to grips with new topics more quickly or identify gaps in their knowledge. Used correctly, AI can thus become a training partner for targeted upskilling, say for expanding skills.
Openness to new ideas is key here. It makes it easier to recognise the positive aspects of the technology and helps ensure that fears of being replaced by new technologies do not hold us back. "Fear prevents us from developing further,” says Ebli-Korbel. Only those who invest in their own development can act with foresight. – dpa
