Will Artificial Intelligence set us free — or turn people into “waste” from progress itself?
The future has already begun — it just still seems unevenly distributed
Do you believe humanoid robots can already work on real production lines?
Yes. BMW has already run pilots with humanoid robots in automotive production environments, including a project in Spartanburg with Figure AI and another in Leipzig with the AEON robot, as part of the idea of “Physical AI”: AI applied to machines capable of acting in the physical world. (bmwgroup.com)
Do you believe driverless cars are already operating in commercial transportation services?
Yes. Waymo operates robotaxi services in several cities, with autonomous vehicles transporting passengers without a human driver behind the wheel. At the same time, recent incidents, such as service disruptions caused by flooding in Texas, show that the technology is real, but still faces practical, climate-related, urban, and regulatory limits. (waymo.com)
Do you believe a person has already controlled a computer using only their thoughts?
Yes. Brain-computer interfaces are already an experimental reality. Neuralink, for example, is developing brain implants with the initial goal of restoring autonomy for people with unmet medical needs; in 2024, it was reported that the company’s first human patient could control a computer cursor through thought. (neuralink.com)
Do you believe “teleportation” already exists?
It depends on what we mean by teleportation. Teleporting people remains science fiction. But quantum teleportation of information has already been demonstrated in real networks: in 2026, Deutsche Telekom and Qunnect announced a quantum teleportation test over a commercial fiber-optic network in Berlin. (telekom.com)
Do you believe AI systems can already write, program, create images, generate videos, compose music, summarize books, analyze data, and simulate complex conversations?
Yes. And perhaps this is the most important point: many technologies that once seemed distant are already entering schools, companies, hospitals, offices, factories, laboratories, and digital platforms. Some are still experimental. Others are already being commercialized. Others work with important limitations. But all of them point in the same direction: the future is not arriving all at once. It is being installed in pieces, product by product, laboratory by laboratory, city by city.
So the question is not only whether these technologies are possible.
The question is:
Who will be prepared to understand them, use them, and question them?
Artificial Intelligence is no longer just a topic for laboratories, science fiction, or large technology companies. In 2026, it is already present in tools for study, work, research, programming, design, music, automation, customer support, data analysis, communication, and content creation.
For students and professionals in training, this changes an important question. It is no longer just:
“Will AI replace jobs?”
The more complete question may be:
“How can I learn to work better with AI in a critical, ethical, intelligent, and conscious way?”
This text is a reflection on the present, but also a provocation about the future. If we continue advancing at the current pace, what kind of society might we find in 2044?
A society where AI expands opportunities, democratizes knowledge, and frees humans from repetitive tasks? Or a society divided between a technological elite integrated into intelligent systems and a mass of displaced, monitored workers treated as disposable?
To think about this, we will go through a brief history of Artificial Intelligence, discuss its socioeconomic, environmental, and political impacts, reflect on the combination of AI and quantum computing, analyze possible future scenarios, and use an analogy with Battle Angel Alita, by Yukito Kishiro.
In Battle Angel Alita, humanity is divided between symbolic spaces of power and exclusion: the Scrapyard, Zalem, and Jeru. This fictional structure helps us think about a future where technology, inequality, consciousness, the body, work, and social control become intertwined.
The idea is not to treat AI as magic, nor as an inevitable threat. The most productive attitude is a combination of curiosity, preparation, critical thinking, and responsibility.
1. A brief history of Artificial Intelligence
The idea of intelligent machines did not begin with ChatGPT, nor with recent generative models.
In 1950, Alan Turing published an important paper on the possibility of machines thinking, which later became associated with the famous “Turing Test.” The central question was provocative: could a machine demonstrate intelligent behavior to the point of being mistaken for a human being in conversation?
In 1956, the Dartmouth workshop helped consolidate the term “Artificial Intelligence” as a research field. From that point on, researchers began imagining systems capable of solving problems, playing games, proving theorems, understanding language, and simulating aspects of human reasoning.
In the following decades, AI went through cycles of enthusiasm and frustration. There were periods of great optimism with systems based on rules, symbolic logic, and manually formalized knowledge. Then came the so-called “AI winters,” periods in which promises exceeded practical results and funding declined.
Later, with increased computational power, the availability of large volumes of data, and advances in machine learning, AI began to grow again. Starting in the 2010s, deep neural networks began transforming areas such as computer vision, speech recognition, machine translation, content recommendation, and image analysis.
An important point in this trajectory was the development of the Transformer architecture, introduced in 2017, which accelerated the progress of modern language models. This architecture helped create systems capable of generating text, summarizing documents, answering questions, writing code, analyzing images, and interacting with users in increasingly natural ways.
The central point is that today’s AI did not appear suddenly. It is the result of decades of research, mistakes, limitations, technical advances, economic disputes, and social change.
Understanding this history helps us avoid two extremes: thinking AI is magic, or thinking it is just a passing trend.
2. Socioeconomic, environmental, and political impacts of AI
When we talk about Artificial Intelligence, it is common to first think about productivity, automation, creativity, and new tools. But AI is not just a “digital” and abstract technology.
It depends on physical infrastructure, electricity, water, minerals, chips, data centers, global supply chains, political decisions, and economic models.
That is why AI’s impact must be analyzed across several dimensions at the same time: economic, social, environmental, geopolitical, and labor-related.
The central question is not only:
“Will AI improve the world?”
A more complete question would be:
“For whom will AI improve the world, who will pay the cost of this transformation, and who may be left behind?”
AI can help students learn better, companies produce more, scientists discover new treatments, governments improve services, and professionals automate repetitive tasks. But it can also increase inequalities, concentrate power, pressure jobs, consume many natural resources, create technological dependence, and generate new political risks.
That is why the discussion about AI needs to go beyond enthusiasm for tools. It needs to consider the entire system that makes this technology possible.
2.1 The economic cost of AI
Modern AI requires extremely expensive infrastructure. Training and operating advanced models depends on GPUs, servers, data centers, energy, cooling, communication networks, training data, specialized engineers, researchers, security teams, and constant maintenance.
At the interface level, an AI tool may seem simple: the user writes a question and receives an answer. But behind that interaction there is a vast chain of physical, technical, and financial costs.
This cost creates an important contradiction. For the end user, some tools may seem cheap, free, or accessible. But building and maintaining the most advanced models requires billions of dollars in investment.
McKinsey estimates that global data center spending could reach around US$7 trillion by 2030, driven by the race to scale AI and computational infrastructure. (mckinsey.com)
This means that only a few companies and a few countries can compete at the frontier of AI.
This concentration of infrastructure may generate a concentration of power. Companies that control models, chips, cloud infrastructure, data, and platforms may influence not only the technology market, but also education, communication, work, research, defense, healthcare, and public services.
In other words: AI may democratize access to certain capabilities, but the infrastructure that sustains that democratization may be concentrated in the hands of a few actors.
This point is fundamental for students and professionals. Using AI is not just about learning a tool. It is about understanding that this tool is part of a much larger economic ecosystem, with commercial interests, strategic disputes, and social consequences.
2.2 The environmental cost of AI
AI also has a significant environmental cost. It depends on data centers that consume electricity, water for cooling, electronic equipment, semiconductors, critical minerals, and telecommunications infrastructure.
The idea that digital technology is “light” or “immaterial” can be misleading. Every AI query, every generated image, every trained model, and every automated system depends on physical machines running somewhere.
Individually, a single question asked to an AI tool may seem insignificant. But when billions of people and companies use these systems daily, the impact accumulates.
Energy consumption is one of the main concerns. The International Energy Agency projects that global data center electricity consumption could almost double, from around 485 TWh in 2025 to 950 TWh in 2030, reaching approximately 3% of global electricity demand; AI-focused data centers are expected to grow even faster during this period. (iea.org)
Water is also an important issue. Some data centers use water directly for cooling. In addition, there is indirect water consumption in electricity generation, depending on the energy mix used.
Another problem is electronic waste. The race for more advanced chips may accelerate equipment replacement, increase demand for critical minerals, and create new challenges for disposal and recycling.
This does not mean that AI is necessarily an enemy of the environment. On the contrary, it can also help in areas such as climate forecasting, power grid optimization, precision agriculture, waste reduction, discovery of new materials, and scientific modeling.
The point is that AI needs to be evaluated with balance: it can be a tool for solving environmental problems, but it can also increase pressure on natural resources if used irresponsibly or without planning.
Some important questions are:
- Does the energy used by data centers come from renewable or fossil sources?
- Does the water used for cooling come from regions facing water scarcity?
- Is the equipment properly recycled?
- Is the expansion of AI increasing or reducing emissions?
- Who suffers the local impacts of installing large data centers?
- How can we balance technological innovation with sustainability?
A technologically mature society should not ask only whether an AI is powerful. It should also ask whether it is efficient, sustainable, auditable, and socially responsible.
2.3 The political and geopolitical cost of AI
AI is also a question of political power.
Countries that dominate semiconductors, data centers, advanced models, cloud infrastructure, data, and cutting-edge research may gain economic, scientific, military, and diplomatic advantages.
This can increase the gap between countries that produce technology and countries that consume technology. In practice, countries that do not develop their own capabilities may become dependent on tools, platforms, and rules defined by others.
This dependence can affect education, national security, research, industry, communication, healthcare, and public administration.
In addition, AI can directly influence political processes. Generative systems can create convincing-looking fake texts, images, videos, and audio. This increases the risk of disinformation, scams, electoral manipulation, automated influence campaigns, and erosion of public trust.
When anyone can create a fake image, a fake voice, or a fake video with a few prompts, society faces a serious problem: how do we know what is real?
This challenge is not only technical. It is also educational, legal, journalistic, and political.
Another central point is regulation. Governments will need to decide how to deal with privacy, copyright, liability for errors, safety, military use, job protection, education, algorithmic transparency, and market concentration.
If regulation is too weak, companies may externalize social, environmental, and labor costs. If it is too rigid, it may block innovation and hinder local development. The challenge will be to create rules that protect society without preventing useful advances.
AI may also change how states operate. Governments can use AI to improve public services, detect fraud, optimize policies, and expand access to information. But they can also use it for surveillance, social control, repression, algorithmic discrimination, and manipulation.
That is why discussing AI is discussing democracy, sovereignty, civil rights, and the distribution of power.
2.4 AI, employment, and the transformation of work
AI’s impact on jobs is one of the most important and sensitive topics.
A common idea is to imagine that AI will simply “steal jobs.” This phrase captures a real concern, but oversimplifies the problem.
In most cases, AI replaces tasks first, not entire professions.
A profession is made up of many tasks. Some are repetitive, predictable, and easy to automate. Others require judgment, experience, empathy, physical presence, negotiation, creativity, legal responsibility, or deep contextual understanding.
The IMF estimates that around 40% of jobs worldwide are exposed to AI; in advanced economies, this number may reach 60%, with some jobs being complemented by AI and others facing reduced demand, wages, or hiring. (imf.org)
Goldman Sachs also estimated that generative AI could expose the equivalent of 300 million full-time jobs to automation. This does not mean that 300 million people will necessarily lose their jobs, but that many tasks performed by humans may be partially automated. (goldmansachs.com)
For example, a lawyer may not be entirely replaced, but tasks such as summarizing contracts, researching case law, or drafting documents may be automated. A programmer may remain necessary, but parts of code writing, testing, and documentation may be done with AI. A teacher may use AI to prepare materials, but still be essential for guiding, motivating, assessing, and supporting students.
The greatest risk is in jobs composed of repetitive, predictable, digital, and easy-to-validate tasks.
At the same time, some professions may become more productive. A professional who uses AI well can produce more, study faster, create prototypes, revise texts, automate workflows, and make decisions with more informational support.
This can generate productivity gains, but it can also reduce the need for labor in certain roles. If a team of ten people can produce the same result with five people using AI, the company may grow, or it may simply cut costs. The outcome depends on economic, social, and political decisions.
Therefore, AI’s impact on employment will not be the same for everyone.
Some people will be augmented by AI.
Some will be displaced by AI.
Some will have their wages pressured.
Some will see their roles change.
Some will create new careers with AI.
And some may be excluded if they do not have access to education, tools, and opportunities.
Perhaps the most important question is not only:
“Will AI end jobs?”
But rather:
“Who will be able to use AI to increase their capabilities, and who will be replaced, monitored, or made precarious by automated systems?”
2.5 What if AI replaces workers across different fields?
The discussion about worker replacement should not be limited to humanoid robots, factories, or physical labor. AI can impact practically any field where tasks are repetitive, standardizable, predictable, or based on information processing.
In fact, one of the major differences between generative AI and previous waves of automation is that it directly affects work considered “intellectual” or “office-based.”
For a long time, automation seemed more associated with industrial machines, assembly lines, and robotics. Now, writing, analysis, customer service, programming, translation, design, research, and decision-support tasks are also beginning to be partially automated.
Replacement can happen in several ways. In some cases, an entire profession may lose demand. In others, the profession continues to exist, but with fewer people, because each professional can produce much more with AI support. The profile of the role may also change: simple tasks disappear, while supervision, validation, strategy, human relationships, and responsibility become more important.
A customer support professional may have part of their work replaced by chatbots and voice agents.
A translator may lose simple direct-translation jobs, but remain necessary for specialized review, cultural localization, and sensitive texts.
An entry-level programmer may have part of their work automated by code generation tools, while experienced programmers focus more on architecture, review, security, and integration.
A designer may use AI to quickly generate visual drafts, but still needs to make aesthetic, strategic, and brand identity decisions.
A financial analyst may have reports automated, but still needs to interpret risks, economic context, and strategic decisions.
A teacher may use AI to create materials, exercises, and lesson plans, but remains essential in human mediation, emotional support, and pedagogical adaptation.
A lawyer may automate research, summaries, and drafts, but remains responsible for interpretation, legal strategy, ethics, and client representation.
A doctor may use AI as diagnostic support, but clinical decisions, communication with patients, and professional responsibility still require human presence.
A journalist may use AI for drafts and research, but investigation, verification, editorial responsibility, and critical reading of reality remain human differentiators.
An administrative professional may automate spreadsheets, reports, emails, scheduling, and internal workflows, reducing the need for repetitive operational tasks.
A researcher may use AI to review literature, organize references, generate hypotheses, and support data analysis, but still needs to master scientific method, interpretation, and validation.
A marketing professional may automate ads, segmentation, and text variations, but still needs to understand human behavior, brand, culture, and strategy.
An accountant may automate entries, reports, and classifications, but will still be needed for tax interpretation, compliance, auditing, and decision-making.
A human resources professional may use AI for screening and analysis, but must be careful with bias, fairness, human context, and ethical responsibility.
In other words, AI’s impact is not uniform. It does not replace all people in the same way. It replaces tasks first, then may reduce the need for certain roles, and only in more extreme cases may make some entire professions less relevant.
For large-scale replacement to happen, some conditions need to be present.
The first condition is that the task must be technically automatable. AI must be able to perform that activity with sufficient quality. Tasks based on text, code, image, audio, classification, triage, recommendation, and standardized analysis are easier to automate than tasks that are highly social, ambiguous, physical, emotional, or morally complex.
The second condition is that automation must be cost-effective. Companies do not automate simply because the technology exists. They automate when the cost of the tool, integration, maintenance, training, and risk is lower than the cost of keeping people doing that task.
The third condition is that quality must be acceptable. In some areas, a small error may be tolerable. In others, an error can cause financial loss, legal damage, risk to life, or reputational harm. The greater the risk, the greater the need for human review.
The fourth condition is that the organization must be able to integrate AI into its workflow. Having a powerful tool is not enough. It is necessary to adapt processes, train teams, protect data, define responsibilities, measure results, and deal with failures.
The fifth condition is that legislation and regulation must allow it. Areas such as healthcare, law, finance, education, security, transportation, and human resources may have legal and ethical limits on full automation.
The sixth condition is that customers, users, and society must accept it. Even if an AI can do something, people may not easily accept being served, assessed, taught, treated, or judged only by automated systems.
The seventh condition is that responsibility must be clear. When an AI makes a mistake, who is accountable? The company? The developer? The user? The professional who accepted the recommendation? In many fields, this remains complex.
That is why the most realistic scenario is not the instant replacement of all workers, but a deep reorganization of work. Professions will be redesigned. Some tasks will disappear. Others will become more valuable.
Entry-level roles may shrink, creating a special challenge for students and young professionals who need their first job to gain experience.
This last point is very important.
If basic tasks are automated, how will new professionals learn?
Many people begin precisely by doing simple activities: reviewing documents, preparing reports, correcting small errors, responding to requests, organizing data, writing drafts, running tests, serving clients, or supporting more experienced professionals.
If these tasks are handed over to AI, companies will need to create new forms of professional training. Otherwise, a “gap” may emerge at the beginning of careers: fewer junior positions, less practical learning, and more demand for experience from people who have not yet had the opportunity to acquire it.
So perhaps the most important question is not only:
“Will AI replace workers?”
But rather:
“Which tasks will be automated, which skills will remain human, and how will we train new professionals in a world where basic tasks may disappear?”
2.6 The apocalyptic scenario: what if AI causes mass unemployment?
An extreme scenario would be one in which AI, automation, and robotics replace a large share of human labor faster than society can create new roles, adapt education, and reorganize the economy.
This scenario does not require AI to be conscious, perfect, or “superintelligent.” It only requires AI to be good enough to perform a large number of tasks more cheaply, quickly, and scalably than humans.
In this scenario, we could see structural unemployment, falling wages in several areas, fewer entry-level jobs, increased competition for work that still requires humans, extreme concentration of wealth in companies that control AI, data, robotics, and infrastructure, rising inequality, a crisis of professional identity, and the devaluation of degrees that failed to adapt to the new context.
We could also see growth in temporary, fragmented, or platform-mediated work; increased surveillance of workers; political conflicts; pressure for universal basic income or new forms of social protection; disputes over who owns the data, the models, and the robots; and the rise of anti-technology movements.
This scenario is not inevitable, but it is possible enough to deserve attention.
The risk does not come only from technology, but from the speed of transition. Societies can adapt better when changes happen with time, education, public policy, social protection, and redistribution of opportunities. The danger lies in a transition that is too fast, too concentrated, and unplanned.
There is also a psychological and cultural risk. Work is not only a source of income. For many people, work is also identity, routine, social recognition, purpose, and belonging.
If many people are displaced at the same time, the crisis may be not only economic, but also social and emotional.
That is why discussing AI and employment is not only discussing efficiency. It is discussing human dignity.
2.7 The most likely scenario: unequal transformation
The most likely scenario in the short and medium term may not be “the end of work,” but an unequal transformation of work.
Some people will use AI to produce more, learn faster, start companies, automate tasks, and compete globally.
Others will be replaced in important parts of their work, have their wages pressured, or become dependent on platforms that control tools, data, and the distribution of opportunities.
Some companies will use AI to improve products, reduce waste, increase quality, and free people from repetitive tasks.
Others will use AI only to cut costs, monitor workers, and transfer risks to individuals.
Some countries will succeed in creating adaptation policies, technological education, local innovation, and social protection.
Others may become only dependent consumers of technologies developed elsewhere.
In other words, AI will not automatically produce a good or bad future. It will amplify the choices society makes.
If used responsibly, it can help improve education, healthcare, science, productivity, and inclusion.
If used only to concentrate profit and power, it can increase inequality, precarize work, and weaken institutions.
3. AI and quantum computing: what could change?
Quantum computing is still at a very different stage from generative AI.
While AI tools are already part of everyday life for millions of people, quantum computers still face major challenges in stability, scale, error correction, and practical application.
Even so, the combination of AI and quantum computing may become one of the most important areas of the coming decades.
The relationship can work in two directions.
The first is AI helping quantum computing. AI models can help design new materials, control quantum systems, correct errors, optimize circuits, and interpret experimental data.
The second is quantum computing helping AI. In the future, quantum computers may accelerate certain types of optimization, molecular simulation, and processing of complex problems. This does not mean that every AI model will be replaced by quantum computing, but that some niches may gain new capabilities.
The most promising areas include optimization, material discovery, computational chemistry, post-quantum cryptography, physical simulations, and some types of quantum machine learning.
However, it is important to avoid exaggeration. Quantum computing is not a magical solution for every problem. Many practical challenges still need to be solved before it has broad everyday impact.
The most balanced conclusion is:
AI is already practical. Quantum computing is still emerging. The combination of the two may be powerful, but it should be treated with caution, without futuristic exaggeration.
4. The analogy with Battle Angel Alita
Science fiction often anticipates dilemmas before they become real problems. It allows us to look at the present through extreme, exaggerated, and symbolic futures.
In Battle Angel Alita, by Yukito Kishiro, we find a world marked by extreme inequality, modified bodies, cyborgs, advanced biotechnology, manipulation of identity, social control, and tension between humanity and technology. The original work is a cyberpunk manga published in the 1990s, with sequels such as Battle Angel Alita: Last Order and Battle Angel Alita: Mars Chronicle. (en.wikipedia.org)
In the work, the hierarchy between the Scrapyard, Zalem, and Jeru is not only geographical. It represents different layers of access, power, and technological control. In older translations, Zalem also appears as Tiphares, and Jeru appears as Ketheres. (en.wikipedia.org)
In the Scrapyard, people live below Zalem and survive from the leftovers of a superior society. It is a space of exclusion, violence, precarity, and extreme adaptation. The city functions as a brutal metaphor for a world where part of humanity lives off the waste of other people’s progress.
In Zalem, the population lives in an elevated city, apparently more advanced, clean, and protected. But this superiority hides a disturbing secret: adults have their organic brains removed and replaced by brain chips, preserving memories and personality. In the mythology of the work, the so-called brain bio-chips perform functions equivalent to the human brain and retain personality and memories. (battleangel.fandom.com)
This detail is essential to our reflection on AI. In Zalem, the problem is not only that technology has advanced. The problem is that it has advanced to the point of altering the very definition of consciousness, identity, and freedom without people fully understanding the structure in which they live.
In Jeru, the discussion becomes even deeper. The space city is associated with systems of control, power, and advanced computation. Among the most important concepts are Melchizedek, Unanimous, and Methuselyzation.
Melchizedek can be understood as a central computational intelligence, linked to governance, infrastructure, and the control of a highly technological society. In the work, Melchizedek is described as a quantum supercomputer linked to Ketheres/Jeru and plays an important role in Last Order. (battleangel.fandom.com)
In a symbolic reading, Melchizedek approaches a question that is also beginning to appear in our world:
What happens when computational systems begin to influence fundamental decisions about work, education, security, access to opportunities, circulation of information, and social organization?
Unanimous expands this question even further. In the work, Unanimous is described as a public order system used by the residents of Ketheres/Jeru, officially associated with access to Melchizedek for information. (battleangel.fandom.com)
As a contemporary metaphor, Unanimous can represent a hyperconnected society where data, algorithms, social networks, reputation systems, digital platforms, and AI begin to influence what people think, desire, buy, study, defend, and believe.
Methuselyzation brings an even heavier dimension: the pursuit of radical life extension, preservation of consciousness, biological manipulation, and overcoming human limits. In the work, Methuselyzation is described as a process that stops aging through personal nanomachines installed in the body. (battleangel.fandom.com)
This idea connects with current debates on biotechnology, brain-computer interfaces, mind uploading, body-part replacement, personalized medicine, medical AI, and transhumanism. We are still far from many of these scenarios, but the philosophical direction already exists: to what extent do we want to use technology to repair, expand, replace, or redesign the human being?
That is why Battle Angel Alita is such a powerful analogy for discussing AI. The work is not only about robots or cyborgs. It is about inequality, social discard, technological elites, modified bodies, replaced brains, control systems, and societies divided between those who dominate the infrastructure and those who survive from its leftovers.
The Scrapyard, Zalem, and Jeru can be read as three possible futures within the same civilization.
The Scrapyard represents those excluded from progress: those who do not control technology, do not have access to the best opportunities, and must survive within the economic and social residues produced by others.
Zalem represents an intermediate or privileged class, apparently elevated, productive, and protected, but dependent on systems it does not fully understand. It is technological comfort accompanied by alienation.
Jeru represents the top of the structure: control over infrastructure, computation, data, biotechnology, politics, and perhaps the very definition of humanity.
This structure helps us think about 2044.
In a future dominated by AI, who will symbolically live in Jeru? Who will have access to the best models, the best data, the best schools, the best networks, the best medical treatments, and the best opportunities?
Who will live in Zalem? Who will be comfortable, productive, connected, and apparently free, yet deeply dependent on invisible systems that decide what they can see, learn, consume, produce, or desire?
And who will be left in the Scrapyard? Who will be discarded by the market, made invisible by platforms, replaced by automation, or reduced to data used to train systems that benefit others?
This analogy is powerful because it raises questions that are also beginning to appear in the real world:
- What does it mean to be human when parts of the body, memory, decision-making, or intelligence can be mediated by technology?
- Who will have access to technological enhancements?
- Will technology be used to free people or control people?
- Will knowledge be democratized or concentrated?
- Will human life become more dignified or more surveilled?
- Will AI be a tool for emancipation or a new layer of inequality?
In our current world, we are still far from scenarios such as the widespread replacement of brains with chips in the style of Battle Angel Alita. But we are already dealing with early versions of these dilemmas: algorithms deciding opportunities, AI influencing education, automated systems evaluating candidates, personal data training models, platforms defining professional visibility, and digital tools expanding or reducing inequality.
Fiction helps us ask an essential ethical question:
It is not enough to ask “can we do it?” We must also ask “should we do it?”, “for whom?”, “with what limits?”, and “with what consequences?”
5. Possible scenarios for the future
5.1 Short term: 2026 to 2030
In the short term, AI should become increasingly common in work and study tools.
Writing assistants, programming assistants, translation tools, summarization tools, data analysis systems, slide generators, task automation, and customer service AI will become increasingly integrated into professional routines.
The main risk will be the illusion of competence. People may submit texts, code, presentations, or analyses generated by AI without understanding the content. This can lead to errors, plagiarism, poor decisions, and intellectual dependence.
The professional differentiator will be knowing how to use AI as a reasoning partner, not as a substitute for thought.
In this scenario, students who learn to use AI critically may gain an advantage. But students who only copy answers may become more fragile, because they will appear productive without developing a real foundation.
5.2 Medium term: 2030 to 2040
In the medium term, some professions may change profoundly.
Repetitive office tasks, customer support, simple content production, document analysis, information triage, and basic programming may become highly automated.
On the other hand, new roles should emerge: AI validation specialists, automated workflow designers, algorithmic auditors, AI-assisted educators, technology ethics professionals, knowledge curators, integration engineers, model security specialists, and professionals who combine AI with specific domains.
The market may value less those who only execute standardized tasks, and value more those who can define problems, validate results, make decisions, coordinate people, and understand context.
Professional education will need to change. It will not be enough to teach ready-made answers. It will be necessary to teach investigation, adaptation, critical thinking, communication, ethics, and the ability to keep learning.
5.3 Long term: after 2040
In the long term, the combination of AI, biotechnology, robotics, brain-computer interfaces, and quantum computing may generate transformations that are difficult to predict.
This is where analogies like Battle Angel Alita become useful, because they force us to think not only about productivity, but also about identity, dignity, inequality, power, and control.
The future will not be defined only by technology. It will be defined by the social, political, economic, and ethical choices we make around it.
Technology can expand human capabilities. But it can also expand human inequalities.
It can free people from repetitive tasks. But it can also create new forms of dependence.
It can democratize knowledge. But it can also concentrate power among those who control infrastructure, data, and models.
That is why thinking about the future of AI is thinking about the future of society itself.
6. How students can prepare for a difficult scenario
Preparing for a difficult scenario does not mean panicking. It means building a more resilient professional foundation.
The best strategy is to develop skills that remain valuable even in a highly automated world.
6.1 Learn how to learn
The most important skill is continuous learning.
Tools, languages, professions, and platforms change quickly. Anyone who depends only on fixed knowledge becomes vulnerable.
Students need to practice reading, research, experimentation, documentation, and constant updating.
Learning how to learn means knowing how to find good sources, compare information, ask better questions, test ideas, and update beliefs when new evidence appears.
In a world where AI can generate quick answers, the advantage is not only in obtaining answers. It is in knowing how to formulate good questions and evaluate whether the answers make sense.
6.2 Develop critical thinking
AI can generate wrong answers with a convincing appearance.
That is why it will become increasingly important to know how to verify sources, compare arguments, identify fallacies, question assumptions, and validate results.
Those who simply accept ready-made answers become fragile. Those who know how to investigate become stronger.
AI should be treated as a support tool, not as a final authority.
A good student needs to ask:
- Is this correct?
- What is the source?
- Is there another interpretation?
- What are the limitations?
- What evidence supports this conclusion?
- What might be missing?
This type of reasoning is similar to the scientific method: observe, formulate hypotheses, test, revise, and improve.
6.3 Combine a domain area with AI
It is not enough to know how to “use AI.” Ideally, AI should be combined with a real field.
- AI + healthcare.
- AI + law.
- AI + administration.
- AI + education.
- AI + engineering.
- AI + communication.
- AI + data science.
- AI + biology.
- AI + finance.
- AI + design.
- AI + languages.
- AI + logistics.
- AI + sustainability.
- AI + public policy.
The competitive advantage will be at the intersection between technical knowledge, human context, and intelligent use of tools.
Someone who only understands AI may generate generic answers. Someone who understands a field and knows how to use AI can solve real problems.
6.4 Learn fundamentals, not only tools
Tools change. Fundamentals last longer.
A programming student should understand logic, data, algorithms, architecture, and security.
A communication student should understand narrative, audience, clarity, persuasion, and responsibility.
An administration student should understand processes, costs, strategy, people, and decision-making.
A science student should understand method, evidence, uncertainty, review, and the limits of data.
A design student should understand visual perception, user experience, culture, brand, and intention.
AI helps most those who already have a foundation to evaluate the result.
Tools can accelerate work, but fundamentals help determine whether the work is correct.
6.5 Develop human skills that are difficult to automate
Some abilities remain highly relevant:
- empathy;
- leadership;
- negotiation;
- ethics;
- contextual creativity;
- communication;
- collaboration;
- conflict resolution;
- strategic vision;
- ability to teach;
- responsibility in sensitive decisions;
- ability to listen;
- ability to deal with ambiguity;
- ability to create trust.
Even if AI helps in these areas, humans will still be needed to deal with social context, moral responsibility, trust-based relationships, and sensitive decisions.
In many situations, the question will not only be “what is the correct answer?”, but “how should this answer be communicated?”, “who will be affected?”, “what risks exist?”, and “which decision is fairest?”
6.6 Learn automation
Instead of competing against automation, students should learn to automate parts of their own work.
This includes using AI to create scripts, organize spreadsheets, generate reports, revise texts, build presentations, answer emails, structure study plans, analyze data, create prototypes, and document processes.
Those who understand automation can stop being replaced by it and become the ones who implement, supervise, or improve it.
This does not mean automating everything without thinking. It means identifying repetitive tasks, creating more efficient workflows, and freeing time for higher-value activities.
6.7 Build a portfolio
In a competitive world, degrees remain important, but practical evidence of ability gains value.
A student can create:
- GitHub projects;
- LinkedIn articles;
- public data analyses;
- presentations;
- small applications;
- technical translations;
- reports;
- educational videos;
- case studies;
- social impact projects;
- teaching materials;
- documented experiments with AI.
A portfolio shows initiative, execution ability, and real learning.
It also reduces dependence on generic resumes. In a market where many people can generate polished texts with AI, concrete evidence of real work will become increasingly important.
6.8 Understand ethics and responsibility
Professionals who know how to use AI safely, privately, and responsibly will have an advantage.
This includes knowing when not to use AI, how to protect data, how to review results, how to avoid plagiarism, how to deal with bias, how to explain limitations, and how to take responsibility for the final work.
The ethical question should not appear only at the end of the process. It should be present from the beginning.
- Am I using data I am allowed to use?
- Am I harming someone?
- Am I being transparent?
- Am I validating the result?
- Am I transferring responsibility to a tool?
- Am I using AI to learn, or just to pretend I know?
A good practical rule is:
The greater the impact of a decision on someone’s life, the greater the level of human validation should be.
6.9 Build an international network
AI can increase global competition, but it also expands global opportunities.
Languages, intercultural communication, and participation in international communities can help students access opportunities outside their city or country.
Open source communities, academic groups, online events, international courses, and collaborative projects can be important paths to learning from people around the world.
In this sense, AI and languages are connected. AI can help with translation and language learning, but being able to communicate directly in another language still opens doors, creates trust, and expands opportunities.
Language remains a tool for breaking territorial barriers, participating in international groups, learning from professionals in other countries, and exposing oneself to greater challenges.
6.10 Prepare emotionally
Rapid technological change generates anxiety. That is why it is also important to develop resilience, adaptability, and emotional maturity.
The professional of the future will need to deal with uncertainty without becoming paralyzed.
This means accepting that it may be necessary to change tools, roles, fields, or strategies more than once throughout a career.
Emotional preparation is not empty optimism. It is the ability to keep learning even when the scenario changes.
7. How to use AI efficiently to learn and work better
The best way to use AI is not simply to ask for ready-made answers.
It is to use AI as part of an iterative process, similar to scientific reasoning.
A good workflow would be:
-
Define the problem clearly.
Before asking AI for help, write down what you want to solve, what the context is, what the constraints are, and what a good answer would look like. -
Ask for a first solution.
Use AI to generate an explanation, plan, code, summary, script, analysis, or list of hypotheses. -
Question the result.
Ask: what are the limitations of this answer? What assumptions were made? What could be wrong? What alternatives exist? -
Verify sources and evidence.
When the topic is technical, scientific, legal, medical, financial, or current, look for reliable sources. AI can be confidently wrong. -
Compare methods.
Ask for different approaches. Compare cost, time, accuracy, risk, and applicability. -
Produce your own synthesis.
The final result must pass through your judgment. AI helps, but responsibility remains human.
This method transforms AI into a tool for active learning. Instead of merely consuming answers, the student learns to formulate better questions, compare hypotheses, revise arguments, and build knowledge.
AI should be used as a reasoning partner, not as a substitute for consciousness.
8. Ten AI tools students can explore
The list below is not a definitive recommendation, but an initial map of useful tools in different areas. Ideally, students should experiment, compare, and choose according to their goals.
It is also important to remember that tools change quickly. Before publishing or presenting this content, it is worth reviewing names, prices, plans, and features.
1. ChatGPT — study, writing, programming, and analysis
It can be used to explain concepts, revise texts, generate ideas, analyze data, create code, study languages, build study plans, and simulate interviews.
It is especially useful when used iteratively: asking, revising, requesting alternatives, testing examples, and verifying results.
2. NotebookLM — source-based study
Useful for studying PDFs, articles, notes, and class materials.
Its differentiator is working more closely with sources provided by the user. It can help students summarize materials, generate review questions, organize ideas, and transform long documents into more accessible explanations.
3. Perplexity — AI-assisted research
A good option for searching answers with references, exploring current topics, and finding sources quickly.
It should be used with critical review, especially for academic, legal, medical, scientific, or professional topics.
4. Claude — reading, writing, and long-document analysis
It can be useful for summarizing long texts, reviewing arguments, structuring essays, and working with complex documents.
It is an interesting tool for those who need to handle long texts, planning, analytical writing, and content review.
5. Gamma — presentation creation
Helps transform ideas into presentations, visual documents, and simple pages.
It can be useful for students who need to present assignments, projects, reports, or ideas in a visually organized way.
Even so, the student should review the structure, arguments, and data. A beautiful presentation does not guarantee a correct presentation.
6. Canva AI — design and visual communication
Useful for posts, presentations, resumes, promotional materials, infographics, and simple visual pieces.
It can help students improve the visual communication of projects, especially when they do not have advanced design experience.
7. Cursor — AI-assisted programming
A code editor with AI assistance features.
It can help plan, write, review, and explain code inside a project. It is useful for those learning programming, but should be used carefully: generating code without understanding it can create dependence and errors that are difficult to fix.
8. GitHub Copilot — programming support
Helps with code autocomplete, generating functions, explaining snippets, and accelerating common development tasks.
It is especially useful for those already learning programming and wanting to practice with review.
Ideally, it should be used as a tutor and accelerator, not as a substitute for learning logic, data structures, tests, and good practices.
9. Zapier or Make — task automation
They allow users to create workflows between applications, such as sending automatic emails, updating spreadsheets, organizing forms, creating notifications, and integrating tools without needing to program everything from scratch.
They are good tools for understanding process automation, something increasingly important in the job market.
10. Suno, Runway, or ElevenLabs — multimedia creation
Tools of this type allow users to explore music, video, synthetic voice, and audiovisual creation.
They are useful for communication, marketing, education, prototyping, and creative projects.
They also raise important ethical questions, especially around authorship, voice use, image use, copyright, and synthetic content creation.
9. The student of the future will not be only an AI user
The student of the future needs to be more than someone who “uses tools.”
They need to understand processes, ask good questions, validate answers, communicate results, and act responsibly.
AI can help people learn faster, but it can also create dangerous shortcuts. It can expand opportunities, but it can also expand inequalities. It can automate tiring tasks, but it can also reduce the autonomy of those who do not understand how it works.
That is why the best posture is neither fear nor worship. It is maturity.
Using AI well means combining curiosity, method, ethics, and practice. It means treating every answer as a hypothesis, not as an absolute truth. It means learning to iterate: ask, test, revise, compare, improve.
In the end, perhaps the most important question is not:
“What will AI be able to do?”
But rather:
“What kind of professional do I want to become in a world where Artificial Intelligence will be increasingly present?”
And perhaps the answer is: a more critical, more creative, more adaptable, more ethical professional, and one more conscious of technology’s impact on society.
The best defense against a difficult scenario is not to ignore AI. It is to learn to use it better than average, understand its limits, develop solid fundamentals, and preserve what is still deeply human: judgment, ethics, creativity, empathy, and responsibility.
In the end, perhaps the future will not simply belong to those who use AI, but to those who can combine AI with consciousness, method, real knowledge, and social commitment.
10. A prediction for 2044
In 2044, perhaps the question will no longer be whether we use AI. Everyone will use AI in some way.
The question will be different:
Who controls the infrastructure? Who understands the systems? Who has autonomy to decide? And who merely obeys the recommendations of machines?
Perhaps some will live as if they were in Jeru: connected to the top of the infrastructure, with access to the best models, the best data, the best schools, the best medical treatments, the best professional networks, and the best protection systems.
Others may live as if they were in Zalem: comfortable, productive, connected, and apparently free, but deeply dependent on systems they do not understand. People who use AI every day, but do not know who defines its rules, its limits, its filters, and its interests.
And many may be pushed into the Scrapyard. Not necessarily a physical city, but a social condition: people discarded by economic systems that begin to treat them as “waste” from progress itself.
In this scenario, the danger is not only that AI replaces human tasks. The danger is that society accepts a new division between the integrated and the discarded, between the optimized and the obsolete, between those who control the machine and those who are controlled by it.
But this future is not closed.
AI does not need to produce a society that treats human beings as disposable parts. It can expand capabilities, democratize knowledge, improve education, accelerate scientific discoveries, support professionals, reduce waste, and help solve real problems.
The difference will lie in the choices we make now: how we educate students, how we regulate technologies, how we distribute opportunities, how we protect workers, how we preserve privacy, how we audit systems, and how we keep human beings at the center of important decisions.
In 2044, the greatest differentiator may not be having a chip in the brain, being connected to a collective intelligence, or living at the top of a technological city.
Perhaps the greatest differentiator will be maintaining the ability to think critically, act ethically, and use technology without completely surrendering one’s humanity.
The future is not written yet.
But it is already being trained.