- Let's talk about AI and the different areas that it has been implemented till now in the whole world, and you can tell me more or less in a interval of time that it will be established and normalized and what will survive in the various and what isn't going forward and is not being very accepted, like for instance in education and use in robotics that's growing. I want to analyze and check all the Proposals of regulation and see also how it will be carried on. And also in terms of the problems in water supplies, the going of the systems, and the worsening of environment in the cities that the data centers are implemented, I needed to do a deep search and see what's in planning for the next three years. And also, well, I know that potentially there will be a lot of shifts in politics, but even so, I think we could expand the analysis till then, including it, and how we're going to get out of those periods in global terms as well, in terms of economy and jobs and all the fears that lower.
- This is one of the most important questions of the decade because AI is not just another technology cycle. It is becoming infrastructure, similar to electricity, the internet, and smartphones. The question is no longer whether AI will be used. The real question is where humans remain central, where AI becomes dominant, and where regulation slows or redirects adoption.
1. Where AI is already becoming normalized
The strongest pattern worldwide is that AI is being accepted when it acts as a copilot, not as a full replacement.
Software and office work
This is probably the fastest area of adoption.
Writing
Coding
Translation
Research
Customer support
Administrative work
Data analysis
Many companies are no longer asking "Should we use AI?" but "How do we integrate it into workflows?"
By 2028, AI assistants will likely be as normal as email and spreadsheets are today.
Healthcare
Medicine
This area is growing rapidly because AI performs well in:
Medical image analysis
Administrative paperwork
Drug discovery
Clinical decision support
However, regulators remain cautious about fully autonomous diagnosis.
The likely outcome:
AI assists doctors
Doctors remain legally responsible
That model is receiving broad acceptance globally.
Robotics and manufacturing
Robotics
This is where many people underestimate AI.
Large language models receive most attention, but physical robotics is advancing steadily.
Factories already use automation extensively.
The next wave is:
Warehouse robots
Logistics robots
Agricultural robots
Elder-care assistants
Semi-autonomous construction systems
The bottleneck is not intelligence anymore.
It is:
Physical reliability
Energy use
Safety
Cost
I expect major normalization between 2026 and 2030.
Education
Education
This area is fascinating because acceptance is mixed.
What is surviving:
Personalized tutoring
Language learning
Study assistance
Administrative support
What is being challenged:
Students submitting AI-generated work
Automated grading without oversight
Many universities are shifting from trying to ban AI to redesigning evaluation systems.
The direction increasingly looks like:
AI allowed, but transparency required.
That is similar to calculators eventually becoming accepted in mathematics.
2. Areas facing resistance
Hiring and recruitment
AI hiring systems face strong scrutiny.
The EU AI Act classifies many employment-related AI uses as "high-risk." Recruitment screening, candidate scoring, and automated assessments are receiving particular regulatory attention.
The reason is simple:
People tolerate AI helping humans.
People dislike AI deciding their future.
Policing and surveillance
This is probably one of the most controversial areas globally.
Especially:
Facial recognition
Predictive policing
Social scoring systems
The European approach is becoming increasingly restrictive regarding certain high-risk or prohibited uses.
AI Act Service Desk +1
Fully autonomous decision-making
Banks, insurers, courts, governments and hospitals face strong pressure to maintain human oversight.
The political consensus emerging in many democracies appears to be:
Human accountability must remain somewhere in the chain.
3. The regulatory landscape through 2028
The biggest regulatory experiment is currently the European Union AI Act.
Timeline
February 2025: prohibited practices and AI literacy obligations began applying.
August 2025: general-purpose AI model obligations began applying.
August 2026: most enforcement starts, including many high-risk systems.
August 2027: remaining regulated-product requirements apply.
AI Act Service Desk +1
The significance is enormous.
Just as the EU's GDPR influenced privacy regulation worldwide, many analysts expect the AI Act to become a global reference point.
What survives regulation?
Likely:
✅ Productivity tools
✅ Educational assistants
✅ Coding assistants
✅ Medical support systems
✅ Industrial robotics
Less likely to remain unrestricted:
⚠ Social scoring
⚠ Emotion detection in sensitive environments
⚠ Fully automated employment decisions
⚠ Certain biometric surveillance systems
�
AI Act Service Desk +1
4. Data centers, water, energy, and environmental stress
This is where things become more complicated.
AI's environmental challenge is not primarily the models.
It is the infrastructure.
Water consumption
Large data centers require cooling.
Many regions already face concerns about:
Groundwater stress
Drought conditions
Competition with agriculture
Urban water supply pressure
Particularly vulnerable areas include parts of:
United States
Ireland
Spain
and some Gulf countries.
The concern is growing because AI workloads consume significantly more computational resources than traditional cloud services.
Electricity demand
This may actually become the larger issue.
Governments and utilities are increasingly planning around:
Nuclear power expansion
Renewable energy expansion
Grid modernization
Localized generation
The next three years will likely see major investment rather than contraction.
Most governments appear unwilling to slow AI growth significantly.
Instead, they are trying to expand infrastructure to support it.
What cities are planning
Emerging strategies include:
Recycled water cooling
Seawater cooling
More efficient chips
Regional distribution of data centers
Building centers near renewable energy sources
The trend is not toward stopping AI.
It is toward making AI infrastructure politically sustainable.
5. Jobs: what probably happens
This is where public fear is highest.
And honestly, there is genuine uncertainty.
Even institutions disagree.
The IMF argues AI could affect around 60% of jobs in advanced economies, depending on how adoption unfolds.
Research increasingly shows:
Productivity gains are real.
Some routine tasks are disappearing.
Entry-level work is especially vulnerable.
New AI-related jobs are also emerging.
The strongest risk is not necessarily mass unemployment.
The stronger risk is transition shock.
Meaning:
Some professions shrink quickly.
New professions emerge unevenly.
Training systems lag behind.
Historically, societies have survived technological revolutions.
The problem is usually the transition period.
6. My expectation for 2026–2028
If current trends continue:
Likely
AI becomes invisible infrastructure.
Most professionals use AI daily.
Robotics expands in logistics and manufacturing.
Governments increase regulation rather than banning AI.
Electricity demand rises significantly.
Data-center environmental conflicts increase locally.
Human oversight remains legally required in sensitive sectors.
Uncertain
Pace of job displacement.
Geopolitical competition between major powers.
Whether productivity gains are distributed broadly or concentrated among large firms.
Whether societies can retrain workers fast enough.
Unlikely
AI disappearing.
Global bans.
A return to pre-AI workflows.
Full replacement of doctors, teachers, judges, or managers in the next three years.
The deeper question underneath all of this is not technological.
It is political and social:
Can societies absorb a productivity revolution without creating a legitimacy crisis around work, inequality, and human purpose?
That is the question governments, companies, economists, and ordinary people are all trying to answer right now—and the answer will probably shape much of the second half of the 2020s.