The steering wheel can be handed over to AI, but life's destination should always be decided by ourselves.
In September 2026, Tesla's Cybercab officially launched limited paid ride service in Austin, Texas. The model's most iconic feature is not yet another iteration of autonomous driving capability, but the complete removal of the traditional steering wheel and the accelerator and brake pedals from the cabin.
Meanwhile, this radically designed vehicle is still under safety review by U.S. regulators, which means its arrival does not equate to the full dawn of the driverless era. Yet it has, at least for the first time, brought a scene that once existed only in science fiction movies into real public view.
Sitting for the first time in a car without a steering wheel, most people instinctively feel uneasy: Where are the controls? How do I intervene in an emergency? Who is responsible when something goes wrong? For over a hundred years, we have settled into a default consensus: whoever sits in a car, the steering wheel must stay in human hands. But what occurred to me is this: we may adapt to this loss of control in far less time than we imagine.
I. Humanity's Underlying Capacity: Not Mastering Execution, but Adapting to Losing Control
The shock brought by a steering-wheel-free car is, in essence, yet another replacement of human execution power by automation technology. And looking back at the history of technology, it is precisely a history of humans continuously ceding standardized execution, with every technological iteration having gone through exactly the same psychological process. When automatic transmissions first became widespread, many believed that manual driving was real driving and that automatics would cost people control of the vehicle; when in-car navigation first spread, there were voices worrying that humans would gradually lose their ability to read roads and judge space.
But the fact is, once a technology's stability and reliability cross the threshold, humans very quickly and voluntarily hand the operations they once performed themselves over to machines — and never want to go back. Today, almost no one refuses an automatic transmission out of fear of "losing driving ability," and very few people give up navigation in order to "train their wayfinding skills." The steering-wheel-free autonomous car is merely the latest node in this same line of evolution. Today we find a car without a steering wheel unthinkable; a few years from now, sitting in a car that still requires manual driving may instead feel like a hassle.
This logic has long been playing out in parallel in the workplace. When AI first helped complete a piece of writing, most people would still double-check it repeatedly, full of doubt; later, when AI began to independently take on complete work modules such as data gathering, report writing, and data analysis, acceptance rose quickly; and now that agents can take on full end-to-end tasks, we have begun to gradually withdraw from the execution stage. This process follows exactly the same law as the transfer of driving rights: as long as the technology is stable and reliable enough, we will naturally cede our execution power.
So I believe the essence of this round of change was never "autonomous driving technology is so advanced," but that humanity is simultaneously adapting to the same thing in two domains: in the car, adapting to the loss of the steering wheel; at work, adapting to the loss of ever more execution power.
II. The Real Deep Crisis: When Execution Is Taken Over, Where Does the Human Value Anchor Rest
Today, the vast majority of discussions about AI stay at the level of "will AI take away my job." Job replacement, unemployment risk, skill obsolescence — these are the surface anxieties brought by technological change. But the more fundamental, sharper question is actually this: if AI really can accomplish more and more of the work, and we really do gradually adapt to this change, then what?
Musk has publicly offered his judgment: with the development of AI and robotics, within the next 10 to 20 years work may become an option rather than a necessity, and even the importance of money may decline along with it. This judgment is undeniably extremely bold, and its realization still requires a great many technical, economic, and social preconditions to be met. But it forces everyone to face, ahead of time, a proposition rarely thought through seriously: if work is no longer the vehicle we must use to prove our own worth, what do people rely on to confirm their meaning?
Since the industrial age, society's value system has in fact been very clear: human value is anchored on execution ability. Writing faster than others, analyzing more accurately than others, operating more steadily than others, being able to complete some complex professional work — these execution-level abilities constitute the core source of self-worth for the vast majority of people. Yet AI is precisely the thing rapidly entering this domain, continuously occupying scenario after scenario at the execution level, with an efficiency far exceeding the pace of human iteration.
So what is truly worth worrying about was never "what new skill AI has learned again," but rather: if we keep defining ourselves by our execution ability, what is left for us when execution ability grows ever less scarce? That, I think, is the real deep cognitive conflict of the AI age — it concerns not the presence or absence of jobs, but the foundation of human value.
III. After the Transfer of Control, People Must Once Again Hold On to Direction and Purpose
The metaphor of "a car without a steering wheel" can, I think, run through the individual transformation of the AI age. An autonomous car can plan its own route, avoid risks, and handle road conditions — it may even drive more smoothly and safely than a human — but at least one thing still requires a human: we still need to tell it where you want to go.
This logic also holds in the relationship between AI and work. AI can solve the "How" problem ever more efficiently, completing the execution stage with speed and quality far beyond human capability. But people must answer two other questions ever more clearly: What — what to do — and, more importantly, Why — why it is worth doing. This is precisely the essential difference between creators and executors: a creator does not need to do everything personally, and can even hand over the vast majority of execution work — but cannot hand over purpose, judgment, choice, and responsibility along with it.
So the truly important capability of the future is increasingly not "can I personally finish writing this report" but "why am I writing this report, and what core problem is it meant to solve"; not "can I make this chart" but "what is worth being presented"; not "can AI do the work for me" but "what do I actually want to create with these unprecedented intelligences." The richer the intelligence becomes, the more important human subjectivity turns out to be. Because as the cost of execution keeps falling, what becomes truly scarce is direction, judgment, choice, and the desire to create.
The steering wheel can be handed over to AI, but life must not be put on autopilot.
We will slowly get used to cars without steering wheels, and we will also slowly get used to large amounts of work being completed automatically by AI. The real question was never whether the technology can do it, but whether we ourselves still know "where we are going" when machines understand "how to get there" ever more clearly.
The real transformation for individuals in the AI age was never about becoming a more skilled AI operator, nor about desperately holding on to the work that AI cannot yet do — it is about finding the human position anew: moving from executor to creator.
From executor to creator: first fulfill the human, then fulfill the AI.
The steering wheel can be handed over to AI, but life's destination should always be decided by ourselves.
