Stop Focusing on “Whether AI Will Destroy Humanity” – First, See Who’s Making the Decisions for You
Hello everyone, I’m your financial journalist and friend economist. Recently, a statistic has been causing quite a stir: some claim that the probability of AI destroying all of humanity within ten years exceeds 10%. This figure has either kept people awake at night, made them believe it’s an exaggeration, or led them to advocate for stopping the development of AI immediately.
But today, I want to present the issue from a different perspective. Let’s not get caught up in whether that 10% is real or not, or debate whether AI is a god or a demon. Instead, let’s consider a more pressing and relevant question: As AI becomes smarter, what place will we humans really have in this new world?
The core idea of this article is quite profound: we’ve always treated AI as an “outsider” – either trying to tame it or fearing it might turn against us. However, the truth is that AI has already become a part of us. The real danger isn’t that machines will suddenly rebel; it’s that, while we enjoy the convenience it brings, we unknowingly hand over the power to make decisions.
Below, I’ll break down this article into five key points to explain in plain language.
---
1. Stop Seeing AI as an “Alien” – It’s a “New Organ” That’s Grown Within Us
Many people think of AI in terms of movies like “Terminator” or “The Matrix”: humans as masters, AI as slaves, or vice versa, as invaders. This “us-versus-them” mindset leads us to either blind optimism (thinking we can control it) or extreme panic (fearing we’ll lose control).
But the author points out that this premise is wrong. AI isn’t a meteorite from outer space; it’s a child of human civilization itself. Fire, writing, machines, electricity – all these were invented by humans. We’ve just started outsourcing parts of our perception, judgment, and action to AI.
It’s like our eyes and brain: the eyes see things, and the brain processes the information. It’s hard to say they’re “hostile.” AI is now like a new organ of our civilization. It indeed has capabilities beyond our expectations, just as a baby learns to walk on its own and parents can’t control every step.
The key point is: Once a powerful technology is created, it’s no longer a static tool; it begins to shape the people who use it. So the question isn’t “how to control this outsider,” but “how to coexist with this new organ.”
---
2. The Question of “Integration” is a Red Herring – Integration Has Already Happened
People still argue about whether we should fully embrace AI or restrict its development. But the author says it’s like asking whether we should breathe air – it’s too late. Integration has already taken place, and it’s much deeper than we realize. Hospitals use AI to analyze medical images, students use it to organize their notes, and companies use it to evaluate resumes. AI is not just a tool; it’s embedded in our systems, interests, and power structures.
You can’t turn back time to a world without AI, just as you can’t separate the internet from our lives. So arguing about integration is pointless; it’s a given fact. The real question is: Where will this integration lead us – to greater strength or weakness?
---
3. Human Value Lies Not in “Being Faster than AI,” but in “Who Bears the Consequences”
We used to comfort ourselves by saying, “AI calculates quickly, but I have emotions; AI processes data, but I have creativity.” As long as AI can’t do something, we’re safe. But the author warns that AI’s capabilities are constantly evolving. What AI can’t do today might be possible tomorrow. If we base our value on what AI can’t do yet, it’s like building a house on sand – it could collapse at any moment.
So where does that leave us? Human value doesn’t lie in being faster than AI; it lies in who bears the consequences of our actions. For example:
- In medicine, AI can predict the success rate of surgeries, but it can’t decide if a patient is willing to take the risks.
- In education, AI can write perfect essays, but it can’t replace the student’s decision about their future.
- In the workplace, algorithms can identify inefficient employees, but they can’t decide who should be fired; the person who loses their job has to face the consequences.
The core logic is: Machines don’t bear the consequences of their decisions; only humans do. Our irreplaceability lies in our ability to be responsible for the results.
---
4. “The Final Signature Is Human” – Don’t Think You’re in Charge
Many think, “As long as I click ‘agree,’ I’m in charge, and AI is just an assistant.” But the author says this is a big illusion. Imagine this process:
1. The system defines the problem (e.g., deciding to lay off employees).
2. The system selects the relevant data (only showing data supporting the layoff).
3. The system generates options (e.g., 10% or 20% layoffs).
4. The system analyzes the risks (indicating 10% as the lowest-risk option).
5. You finally click “approve.”
In this process, the decision is made long before you sign. Your approval is just a formality. You’re not really participating in the judgment; you’re just executing the system’s will.
The real power lies in the ability to say “no”: Can you question the definition of the problem? Can you identify missing information? Can you propose a different solution? Can you request a rethink? If something goes wrong, the responsibility can’t be shifted to AI. Humans must bear the consequences.
---
5. Beware of the “Slowly Drowning Frog” Effect: Progress Doesn’t Necessarily Mean Strengthening
The author mentions a concept called “vector coupling,” which refers to how the combination of AI and humans can either enhance or replace each other. It depends on how we integrate them:
- Positive coupling: Experienced doctors use AI for more accurate diagnoses; lonely people use AI to chat and still maintain social skills.
- Negative coupling: Young doctors become overly dependent on AI, losing their judgment; children use AI to do homework, losing the opportunity to think for themselves; employees rely on algorithms, losing control over their work.
The worst outcome is not that machines take over; it’s that humans gradually withdraw from the decision-making process. Even if things get better (e.g., more efficiency, more output), it doesn’t mean we’re stronger. Convenience doesn’t necessarily make systems more just.
What constitutes “positive coupling”?
- Humans retain the right to say “no.”
- Our judgment doesn’t weaken due to technology.
- Machines do the work, but we understand the principles, can correct mistakes, and can rebuild.
- Responsibilities are clear and not blurred by technology.
If the system’s output improves, but the people doing the work become less empowered, and the system becomes smarter while we lose our voice, that’s not progress; it’s a shift in power.
---
Summary: We Need to Reach for the Stars, but Not Lose Control of the Ship
Back to that 10% statistic – it reminds us of the risks. But a more urgent and real crisis is that we’re gradually losing the ability to define problems, make decisions, and bear the consequences. We don’t need to shut down AI, nor do we need to hand over control. What we need to explore is a new form of civilization where:
- Machines expand our capabilities.
- Humans extend our sense of responsibility.
What’s needed for a journey to the stars is not just a faster “ship” (AI); it’s a sober captain (human beings). Only when our abilities and responsibilities grow together can we truly reach the stars. Otherwise, we’re like people lost in a super-intelligent spacecraft, drifting aimlessly through the universe.
This is what everyone should think about: while enjoying the convenience of AI, do we still retain the ability to think independently and take responsibility for our actions?