How Should Ordinary People Live in the Age of AI?

In 1954, Heidegger delivered a lecture titled The Question Concerning Technology, in which he made a startling claim: "The essence of technology has nothing to do with technology." What he meant was that technology is not merely a tool, not simply a means to an end. Technology is a force that causes the world to reveal itself to us in a particular way. Every paradigm shift in technology changes not only what we can do, but how we see the world — and how we understand ourselves.

Seventy years later, with the true arrival of large language models (LLMs), Heidegger's question has become more urgent than ever. The AI assistant on your phone drafts your emails. AI image generators make design barriers vanish. AI coding assistants let non-programmers write working software. All of this is happening far faster than most people expected.

And riding alongside that speed is a pervasive anxiety. Geoffrey Hinton, the Turing Award-winning godfather of deep learning, publicly stated after leaving Google that he regrets his life's work, because he now believes AI may pose an existential threat to humanity. Elon Musk has called AI the most dangerous technology humans have ever invented. Stephen Hawking warned before his death that the emergence of strong AI could be the best thing — or the last thing — to happen to human civilization.

But most of us are not Hinton, not Musk, not Hawking. We don't have the power to steer AI's trajectory, the authority to set boundaries on this technological revolution, or even enough information to judge which of the smartest people on earth is right. We are ordinary people — with ordinary jobs to do, ordinary families to support, and ordinary lives to live.

Whether AI will destroy humanity or replace every job — these grand anxieties are not the questions you need to answer first. The question you actually need to answer is smaller and more specific: In this AI era that has already arrived, how should I, this particular person, live?

As an ordinary person, these are the questions I've thought about most over the past year. This article won't teach you which tools to use, and it won't give you a productivity checklist. It simply wants to share my reflections: in this era where AI has truly arrived, what kind of people should we become, and with what posture should we coexist with this new presence?

1. AI Leverage Amplifies Everything — Including the Gap

Every major technological revolution in history has failed to eliminate the gap between people. In the short term, it has always widened it. The printing press gave literate people an exponential information advantage over those who couldn't read. The spread of electricity and industrial machinery allowed those who could harness new production tools to rapidly accumulate wealth, while those who couldn't keep pace were left behind. The steam engine liberated physical labor, but the first people to master it were not the hardest-working laborers.

AI follows exactly the same logic. AI's benefit to high-skilled knowledge workers far exceeds its replacement of low-skilled repetitive work. A lawyer using AI can complete in one day what previously took a week of document analysis. A researcher using AI can simultaneously track the information volume that previously required five people.

MIT professor and The Second Machine Age author Erik Brynjolfsson has pointed out: Digital technologies and AI enormously reward those with advanced cognitive skills, not only widening the gap but creating winner-take-all polarized markets.

OpenAI CEO Sam Altman has said directly: "AI will be the greatest force for economic empowerment... but it will also accelerate the trend of inequality."

Forget the distinctions between rich and poor, white collar and blue collar. The future world has only two kinds of people: those with AI leverage, and those without it.

AI is not an equalizer — it's a multiplier. Whatever you bring to it, it amplifies. Marx said the means of production determine the relations of production. In the AI era, cognitive capability itself is the most important means of production. Those who possess quality cognitive frameworks effectively own the best production tools of this era.

Acknowledging that the gap will widen is not about manufacturing anxiety. It's about helping ordinary people see clearly: entering the arena now matters more than ever. AI tools are penetrating virtually every knowledge work domain at extraordinary speed. Whether you choose to seriously learn how to collaborate with AI now, or choose to wait — that decision itself is already widening the gap.

2. Use AI to Expand Your Strengths, Not Patch Your Weaknesses

After entering the arena, many people make the same mistake: using AI to chase areas where they're weak. They see AI can help write code, so they try to learn AI-assisted programming despite having zero feel for technology. They see AI can do data analysis, so they try to use AI to compensate for their quantitative shortcomings.

This is the most common — and most expensive — mistake ordinary people make in the AI era.

In the past, to survive in society, we often had to spend enormous energy "shoring up weaknesses" because the cost of outsourcing was too high. But AI has fundamentally changed this economic premise.

Professional athletes never spend precious training time on their weaknesses — they outsource non-core work to their team and bet all resources unreservedly on their core competitive advantage. In the AI era, you should manage yourself like running a top-tier company: outsource all the peripheral tasks you're not good at — basic coding, data processing, copywriting polish — to AI, your cheap and omnipotent super assistant. Then invest 100% of the time and energy you save into your absolute strengths.

Why? Because under AI leverage, mediocre generalists will rapidly depreciate, while extreme single-point breakthroughs will command unprecedented premiums. A person with 10x depth of cognition in a vertical domain, supplemented by AI to fill tool gaps, can easily defeat dozens of "passing grade in everything" generalists.

What exactly is your strength? It's usually not a single easily-standardized professional skill. It's your deep insight accumulated over years in a particular industry, the sharp judgment born from unique life experiences, the trust and relationships built over time in specific circles, or even your unique taste. These moats are things no advanced language model can replicate by crawling data.

Your weaknesses can be freely outsourced to machines. But your core strengths — only you can deepen those.

But deepening your strengths has a prerequisite more important than any tool: protecting your attention.

Herbert Simon recognized an iron law back in the 1970s: "A wealth of information creates a poverty of attention." Half a century later, this insight carries far more weight than even he could have imagined in his time.

Information in the AI era is not just abundant — it's approaching infinite. AI-generated content flooding in daily, algorithmically-pushed personalized feeds, notifications from countless tools — all compose an endless battle for your attention. The time each person can spend in deep thought each day is fixed and finite. While AI helps you handle low-value tasks, it simultaneously bombards you with endless content. How you allocate your attention determines whether AI is your tool or your noise.

How deep someone can go in their area of strength depends not only on how good their tools are, but on how much uninterrupted focus time they can protect for that domain. Deep work requires active filtering. Focus requires conscious choice.

3. Capture Asymmetric Cognitive Dividends

On the input side, AI has fundamentally changed how ordinary people access information and knowledge. This change is not just about speed — it's about breaking through solid civilizational barriers.

On one hand, there's unprecedented speed and efficiency. Previously, grasping the landscape of a new field required weeks of reading redundant, scattered materials, painfully distilling the main thread from noisy information. Now, AI can read and organize massive amounts of information for you in minutes — and more importantly, it can tailor that information to your existing level of understanding. It's no longer a rigid search engine, but your personal research assistant, helping you quickly penetrate the surface of information and reach the core structure.

On the other hand — and this is the deeper dimension — large language models have for the first time made breaking language barriers cheap.

Language is not just a communication tool. It's the vehicle of thought and the boundary of civilization. What languages you can read determines which information universe you live in. Despite Chinese being our native language, the most cutting-edge academic research, industry analysis, and geopolitical insights are still overwhelmingly in English. Japanese engineering literature, German sociological resources — these are virtually invisible on the Chinese internet.

Previously, foreign language ability was the barrier to first-hand knowledge. Without foreign languages, you could only access translated and filtered second-hand information, bearing inevitable information loss and perspective filtering.

Large language models have made breaking this barrier cheap for the first time. You can now paste a Japanese research report directly to AI and have it not just translate the words, but help you understand the industry context, terminology, and implicit assumptions. This isn't clunky machine translation — it's genuine cross-language understanding.

Entering another language is entering another way of seeing the world. The differences between languages reflect different civilizational understandings of the world. AI has expanded your language boundaries, which in the truest sense expands your world boundaries.

This is one of the most overlooked dividends for ordinary people in this era. Information arbitrage opportunities never come from speed — they come from reaching places others can't see.

4. Say What You Mean — Clearly

Language is not just a tool for reading. It's also your only interface for interacting with AI. What a large language model can give you depends entirely on what you can express to it. This points to a severely underestimated competitive advantage: the ability to say what you mean clearly.

Many people get mediocre results from AI not because the model isn't good enough, but because they themselves don't know what they want. They give AI a vague instruction, get a vague result, then complain that AI isn't smart enough. But the problem usually lies with the person asking, not the system being asked.

Wittgenstein said: "The limits of my language mean the limits of my world" (Die Grenzen meiner Sprache bedeuten die Grenzen meiner Welt). If you can describe a problem precisely in language, it means you truly understand that problem. Conversely, if you can't clearly describe your needs, it most likely means your own thinking hasn't been sorted out.

Pascal said: "If I had more time, I would have written a shorter letter." Precision requires deep understanding. Brevity requires thorough thinking. The prerequisite for saying things clearly is deep reflection on yourself and the problem — What is your goal? What are your constraints? Do you truly care about the result or the process? What can be compromised, and what absolutely cannot?

No AI can answer these questions for you. They can only be slowly clarified by yourself, through continuous self-examination.

This is also why writing has become more important in the AI era, not less. Writing isn't just output — it's the developing fluid of thought. When you try to write a vague idea clearly, you're forcing yourself to truly think about it. Those who can write clearly are often those who can think clearly. And those who can think clearly can maximize the value AI delivers.

Treating clear expression as deliberate practice — this may be the highest-ROI exercise for ordinary people in the AI era.

5. AI Is Your Second Brain, Not Your First

Clear expression is built on active thinking, which brings me to a more fundamental principle: always keep yourself in the loop.

In AI safety, there's a term called "Human in the Loop" (HITL) — keeping human participation and oversight in AI decision-making processes. But I want to borrow this concept for a more everyday context: in every aspect of your AI usage, maintain your active engagement.

Total automation is a trap. GPS navigation has caused many people to lose basic spatial awareness — without GPS, they're completely lost. Calculators have stripped a generation of mental arithmetic instincts. These are localized, relatively mild forms of capability atrophy. But if you outsource thinking itself to AI, what atrophies is your most fundamental capability.

Letting AI think for you is a new form of immaturity. The right posture is: use AI to expand your information reach, use AI for initial synthesis and organization, use AI to test whether your ideas have obvious holes — but core judgments, value prioritization, and final decisions must be your own. AI is the best assistant in your decision chain, but it cannot be the decision-maker itself.

At the same time, AI can be an excellent second brain, but it cannot replace your first brain. There's a classic concept in knowledge management: systematically organizing notes, ideas, and materials into an external knowledge base you can access anytime. AI has supercharged this concept — you can feed it massive amounts of material and have it connect, categorize, distill, even generate first drafts.

But there's a subtle trap here: storing things or having AI process them is easy. Truly internalizing them into your own mind is another matter entirely. Taking notes doesn't equal understanding. Having AI summarize doesn't equal digesting. The internal cognitive framework that can only be formed through extensive reading, repeated thinking, and continuous refinement — there's no shortcut around it.

The stronger this framework, the more valuable insights you can extract from the fragments AI gives you. The weaker this framework, the more AI gives you is just noise that leaves you overwhelmed.

Building your own knowledge framework is more important in the AI era than ever before — precisely because external information is more abundant than ever before.

6. Large Language Models Are Just a Word Game

At its core, a large language model is a probability-based word prediction system. Given a context, it predicts the most appropriate next word, then the next, cycling on and on, composing the fluent responses you see. It can simulate empathy. It can generate words that sound warm and caring. But it does not possess emotions. It does not understand you. It does not care about outcomes.

As AI becomes increasingly human-like, more and more people are treating it as a confidant, an emotional outlet, even a substitute for intimate relationships. This is a dangerous direction — and it's being reinforced by commercial interests, because creating emotional attachment to AI is one of the best business models.

Bertrand Russell said: love specific people, not humanity in the abstract. Abstract love is cheap, even escapist. The care AI gives you is infinite, indiscriminate, and will never reject you — but precisely because of this, it's false.

Real human relationships are real precisely because they have friction, rejection, and imperfection. The moments that hurt you, the cracks that need active repair, the situations where you must learn to compromise — these are things AI can never give you, and these are exactly what make you a truly social being, what make you continue growing through collision with others.

Counter-intuitively, meaning often hides in the friction that technology considers inefficient. Handwriting notes is slow, but the touch of pen on paper creates irreplaceable memory imprints. Walking to the library is far less efficient than querying AI directly, but the aimless associations along the way are often the breeding ground for inspiration. Arguing passionately with friends is far more alive than having AI generate a balanced, objective summary.

In an era when AI can instantly generate everything, ordinary people who actively preserve these inefficient rituals are not only maintaining the warmth and depth of our interaction with the real world — they're continuously reminding themselves of what it means to be human.

Actively choosing to preserve certain inefficiencies is no longer a sign of being behind the times. In an era of extreme acceleration, it's a form of mature self-defense — and a quiet but powerful response to the age of AI.

7. Stay Alive — In Every Sense of the Word

Finally, there's one more thing I want to say. It might sound overly simple, but I increasingly feel it's more decisive than any grand advice: Stay alive. Stay alive as long as you possibly can.

No one knows where AI will ultimately lead humanity. On one hand, AlphaFold has cracked the protein folding problem that puzzled biologists for half a century. AI-assisted genomics is bringing personalized cancer treatment into clinical reality. Brain-computer interfaces are offering hope for neurodegenerative diseases. Conquering cancer and slowing aging are moving from science fiction to reality at visible speed. You will very likely witness a fundamental leap in life sciences.

On the other hand, within Fermi's Paradox lies the famous "Great Filter Theory": sufficiently advanced civilizations tend to self-destruct at some technological threshold — which may be exactly why the universe is so vast yet we remain alone. Is AI humanity's Great Filter? No one knows yet.

But the prerequisite for all of this is: you need to be alive to witness the answer.

In this era of extreme uncertainty, physical health is the most stable and most underestimated "foundation asset" in an ordinary person's portfolio. Sleeping well, exercising consistently, eating healthily — these seemingly unrelated-to-cutting-edge-technology daily habits actually carry genuine strategic significance in the AI era. Because whether technology's wild ride takes us to utopia or the abyss, you need a body strong enough to perceive it, participate in it, and when necessary, influence it.

Final Words

Back to the original question: What has AI changed? It has changed tools, efficiency, the flow of information, and is changing the distribution logic of power and knowledge. But one thing it has not changed, and cannot change: you need to decide for yourself what kind of person you want to become. You need to answer for yourself where your life should point.

Socrates said the unexamined life is not worth living. The greatest risk of the AI era is not being replaced by some algorithm — it's losing self-examination amid infinite convenience. When everything can be outsourced, automated, and optimized, the most fundamental questions of human civilization resurface with even greater clarity: What is efficiency for? Where is this speed taking us? We're accelerating — but do we know our destination?

The more powerful the tools, the more important the direction. AI can help you run faster, but it cannot decide where you're going. A person sprinting at full speed in the wrong direction, with the best AI in the world, is just going off course faster.

Find your strengths. Guard your attention. Say what you mean clearly. Keep yourself in the loop. Build your own knowledge framework. Spend time with real people. Preserve the inefficiencies worth preserving. And then — stay alive.

The most fundamental question of human civilization has never been about efficiency. It has always been about meaning. In the age of AI, that question hasn't disappeared — it's only become more urgent. And the answer can only ever come from you.