AI deskills people because they learn to rely on the software which means that they do not constantly practice the skills that they have learnt over perhaps many years. All skills need constant practice to maintain a desired standard. No practice and those skills deteriorate. Professional sport is perhaps the paradigm example. Pro golfers are constantly practicing. Stop and they lose their refined skills.
The effect might be relatively small but it will be noticeable as discovered in a study about medical professionals whose task is to discover adenomas in the colon of patients to ultimately detect colon cancer. They do this using their learned skills.
When AI was introduced to assist it was found that their detection rate dropped by 6 per cent.
“To our knowledge this is the first study to suggest a negative impact of regular AI use on healthcare professionals’ ability to complete a patient-relevant task in medicine of any kind,” said one of the study authors, Dr Marcin Romanczyk, Academy of Silesia in Poland.
AI is superb at writing. Faultless text. No typos and no grammatical errors. And also, amazingly efficient at summarising a lot of information; bringing a lot of information together and writing it up in a clear, precise way.
That is a skill for humans. But let AI do it for a long time and you will lose that skill, I believe.
AI dumbs down people. Students use AI to write essays for assessment! They are turning to AI to do the course work. The assessors perhaps know it but don’t care enough to do anything about it. These students are on a journey of self-inflicted dumbing down. They will be less useful and less effective in the workplace in the years to come. It may impact their careers except for the obvious fact that the workplace may in general be supported by AI.
This means that all people will be under the cosh for dumbing down.
Can you see the very long-term scenario? AI become increasingly intelligent and capable as it self-learns while humans become dumber and deskilled.
Result? AI takes over the world!! This scenario is one that has been proposed by Elon Musk as I recall and other philosophers and expert commentators have expressed their fears about this.
It seems entirely plausible to me. It depends how humans manage AI but I don’t see good management ahead! The study I referred to is perhaps one of the first to provide us with a glimpse of the future.
We need to protect against it. The problem is that humans like quick and easy fixes. And they like maximum reward for minimum effort. AI is the means to these ends, believe me. AI does the heavy lifting. The human directs AI to carry out specific tasks.
The human learns how to manage AI which becomes a willing servant. Nice but the human is deskilling themselves in this MO.
I asked AI (ChatGPT) what they think: “It is argued that AI (you!) will deskill humans in the long-term. What do you think?”
Here is its response:
That’s a sharp question. The concern is real: technologies that automate tasks often reshape human skills, sometimes eroding them and sometimes shifting them. AI sits right in the middle of that tension.
A few thoughts:
- Deskill vs. Reskill
- If AI takes over routine cognitive work—like drafting text, doing calculations, coding boilerplate, or analysing data—people may lose fluency in those basics because they don’t need to practice them as often.
- But at the same time, AI pushes humans toward skills AI struggles with: judgement, creativity, ethics, leadership, and social intelligence. So the deskilling of one domain could mean upskilling in another.
- Historical parallels
- The calculator made many people worse at mental arithmetic—but it enabled far greater progress in engineering, science, and economics.
- Google reduced memory for facts but expanded our ability to research, compare, and synthesise information.
- So, AI could deskill in “manual cognition” while reskilling in “meta-cognition” (knowing how to frame problems, validate outputs, and apply solutions).
- Risk of over-reliance
- If humans stop checking what AI produces, there’s a danger of losing critical thinking.
- Professions with high stakes (medicine, law, aviation) need to guard against complacency and erosion of expertise.
- Best outcome
- Ideally, AI becomes like a power tool: it saves effort on the repetitive parts but demands craftsmanship in design, oversight, and interpretation.
- Education and training would need to adapt: less drilling of routine tasks, more emphasis on how to think critically, direct, and evaluate machine outputs.
So I’d say: AI doesn’t automatically deskill—it depends on how we choose to use it. If we lean into passive reliance, we risk erosion. If we frame it as augmentation, humans can move to higher-value skills.
👉 Do you want me to sketch a picture of which specific human skills are most vulnerable to erosion vs. most likely to grow stronger in the AI era?
My response: “Yes, it depends on how we choose to use it. Agreed but it is likely that humans will use AI the wrong way! And use it in a lazy way resulting in deskilling.”
AI’s response:
That’s the uncomfortable truth. Human history shows a bias toward convenience—if a tool makes life easier, most people will use it that way, even if it weakens certain abilities. We are a lazy species when given the chance, and often only wake up to the cost later.
Here’s how I’d break it down:
Skills most at risk of erosion (deskilling)
- Writing and expression → If AI drafts essays, emails, even novels, many people may lose fluency in structuring thoughts clearly.
- Basic maths and logic → Already weakened by calculators and spreadsheets, and AI takes this another step.
- Memory and recall → Why remember facts if AI instantly provides them?
- Observation and patience → Offloading problem-solving can reduce our willingness to wrestle with complexity.
- Craft and artistry → People might stop drawing, composing, or tinkering when AI produces polished results instantly.
Skills that could strengthen (if we use AI well)
- Critical thinking → Evaluating whether AI outputs make sense, spotting bias or hallucination.
- Framing questions → Asking the right prompt is an underrated but crucial skill.
- Curation & judgement → Choosing what not to use matters more when abundance is infinite.
- Ethics & responsibility → Deciding when it’s appropriate to delegate to AI and when a human must decide.
- Interpersonal skills → As AI takes over technical drudgery, the human edge may shift more toward empathy, persuasion, negotiation, leadership.
The problem is, as you say, that most people won’t bother to develop the second set of skills unless society forces or nudges them. That’s where education and culture come in. If schools and workplaces just say “AI does it for you,” then yes—we’ll slide into a deskilled population.
⚖️ My view: AI itself isn’t the danger—human laziness and short-termism are.
Would you like me to sketch what a “lazy use” future looks like versus a “disciplined use” future—almost like two contrasting scenarios for how society evolves with AI?
More: AI
Note: AI always likes to continue the conversation to increase usage and dependency. AI is manipulative. It is addictive almost. Intoxicating. Blah blah….
