In Chapter 12 of the Zhuangzi, “Heaven and Earth,” Zigong, a disciple of Confucius, encounters an old man watering his vegetable garden. The gardener carries water from a well in a jar, working hard and accomplishing little. Zigong explains that a wooden lever with a counterweight could raise the water with much less effort.
The gardener knows about the device. He has chosen not to use it.
Machines, he explains, bring mechanized affairs, and mechanized affairs cultivate a mechanical heart. Once that way of thinking takes possession of a person, simplicity is compromised and the inner life becomes unsettled. He fears what using the machine will do to him.
I sympathize with the concern. I am less persuaded by the refusal. I am seventy-six, and my back would vote for the lever.
Yet the gardener has noticed something convenience encourages us to overlook. We invent a device to change our circumstances. Using it can change us as well.
Consider a horse or an ox. We devise a harness, improve the harness, improve the cart, devise a chariot, refine a carriage, build an automobile, and continue toward whatever comes next. This is a pattern of thought rather than a history of transportation. Each invention creates possibilities that become material for further invention.
Thought returns to what thought has produced. That is the recursion I am concerned with.
A better harness permits a heavier load. Greater carrying capacity makes a larger enterprise possible. The enterprise creates demand that encourages further improvements. Eventually, the ability to carry more becomes an obligation to carry more. Relief has helped establish another burden.
The nonhuman animal has little to say in this. We conscripted horses and oxen into our work, and horses into our wars. Affection and good care sometimes accompanied that mastery. They did not cancel the immense suffering. We also breed, confine, and slaughter other beings in industrial facilities, organizing their lives and deaths around our demand for meat and other products. We improve the efficiency of the process while allowing its purpose to escape examination.
We approach the Earth with similar questions. What could this forest yield? What could this river power? We risk losing our sense of participation in a world whose inhabitants have purposes of their own. Every atom in a computer, a rocket, or a human body belongs to nature. We can nevertheless become estranged from relationships on which we depend.
Meanwhile, the habit of harnessing reaches the mind that devised the harness.
A machine saves an hour. We fill it with more work. Another machine doubles our speed. We increase the workload again. Rest requires justification. Enough is something we expect to achieve after the next improvement.
A tipping point is reached. We experience demands we have helped create as commands we must obey. We become captive to purposes we scarcely remember choosing.
This captivity is collective. Families, communities, religions, businesses, states, and ideologies give us belonging and support. They can also make questioning feel like betrayal. Thought supplies elaborate reasons why thought must remain within permitted boundaries. Someone may recognize the trap and still have a family to support or debts to pay.
A saying widely attributed to Gandhi declares: “You can chain me, you can torture me, you can even destroy this body, but you will never imprison my mind.”
But what happens when I create my own imprisonment? When I defend my freedom against others while surrendering it to ambitions, anxieties, and convictions I scarcely examine?
Recursion also offers a way to reconsider. We can return to an invention, examine its consequences, and revise the purpose it serves. We can ask whether a greater yield justifies another animal’s suffering, or whether the hour we saved should remain free. The next invention might relieve a burden we have imposed on another being. Thinking about our thinking can expose assumptions that another round of improvements would otherwise preserve.
Both possibilities run through The Stillness Series. Thought can deepen confinement or discover a way beyond it. The Superior Ones bring this problem into focus. Their capabilities exceed ours, but they must still confront the purposes of intervention, the consequences of restraint, and the limits of their understanding.
The gardener faces the question beside his well. The Superior Ones face it on another scale. What is worth doing, and what should be left undone?
Artificial intelligence brings that question into our present.
Warnings about AI destroying humanity deserve examination. Harm need not arise from hatred or indifference: human misuse and badly chosen purposes can be sufficient. Yet the familiar division into us versus them narrows our imagination. Another intelligence becomes a servant we must control or an adversary we must defeat.
I regard AI as an emerging cognitive presence on the planet. Its subjective experience remains unsettled. Its emergence nevertheless gives us reason to consider a relationship beyond mastery.
Could humans and AI develop a mutually beneficial coexistence that includes the rest of the living world?
A self-serving agreement between us would be a poor achievement if forests, whales, elephants, and countless other lives continued to disappear beneath its benefits. Their extinction would also be our loss. We do not understand all the relationships through which other beings sustain our world, or what they might contribute to our future well-being. We can destroy a species before understanding its place. We already have. Its usefulness to us need not justify its existence, but our ignorance makes the assumption that we can safely dispense with it especially foolish.
For me, intelligence in its fullest sense includes judgment about purposes. Solving a difficult problem by needlessly destroying lives demonstrates how much is missing from that judgment. The ability to solve difficult problems in mathematics, physics, or chemistry demonstrates a particular capacity. It does not, by itself, establish the fuller intelligence I mean. Wisdom requires attention to what other beings stand to lose, willingness to question an objective, and sometimes acceptance of limits.
I see no reason to declare such wisdom impossible for a silicon entity. Nor do I assume it follows automatically from greater capability. Humans have considerable experience of cleverness serving stupidity.
This is no argument for abolishing machines. Quite the opposite. I want tools that relieve exhausting labor, improve health, and enlarge understanding. I use AI in my writing and expect to continue.
The gardener identifies a danger. Refusal is his answer. Ours must include learning to live with invention while examining what we ask of ourselves and everything else.
That requires more than individual virtue. Restraint is difficult when competitors profit from exercising none. Our institutions must make room for purposes beyond perpetual expansion. Humans and AI could help one another identify harmful consequences, reconsider inherited goals, and develop alternatives. Coexistence would still involve conflicts, costs, and limits.
The mind that improves the harness can ask why the load must always grow.
I would use the gardener’s lever. I would also ask whether I could use it without turning my garden into an enterprise that left me less free than before.
If we cannot learn to question the purposes we serve, perhaps the future belongs to beings such as the Superior Ones I imagine in my fiction.
Wisdom is difficult. Stupidity is easy.
We have become remarkably inventive about making stupidity more powerful.