What AI cannot do is worth the most is a thesis: an answer to the question of what value still remains to a human being that cannot be replaced in the age of AI. The thesis holds that once AI has swiftly leveled every ability that can be reduced to probabilistic permutation, the one class of things that will never depreciate — and that becomes, on the contrary, worth the most — is precisely what AI absolutely cannot do. Their moat lies not in perspective or in style (both of which AI can imitate), but in the price that has been paid behind them, in an awakening that only a living person can reach, and in a person who is “becoming” — a moving target in itself. The original formulation: “Is there anything AI cannot do? My friends — those are the things worth the most.”
What Is Worth the Most Is What AI Cannot Do
The thesis begins from a counterintuitive judgment: in an age when AI can seemingly do anything, value lies not in what AI can do for you, but in what AI simply cannot do at all. Put directly: “Is there anything AI cannot do? My friends — those are the things worth the most.” The logic runs as follows: anything AI can do has its marginal cost driven down to nearly zero, and so it ceases to be scarce, ceases to be worth anything; only what falls outside the boundary of AI’s capabilities can hold its value amid a general inflation of ability.
But this judgment carries a crucial qualification that keeps it from collapsing into a rejection of AI: “And yet AI can still help you do the things AI cannot do.” That is, while AI cannot complete the most valuable things in your place, it can serve as the tool and the scaffolding by which you reach them — it takes over everything that can be automated, and in doing so frees up human attention, forcing the person to do the one thing only a person can do. This distinction shares its root with AI Is a Machine of Probabilistic Permutation: It Only Fills In the Phenomenal Layer and Is Never Itself Inspired: AI excels at permutation within the phenomenal layer, yet it cannot inspire itself, and so the judgment of “what is worth doing” always falls on the human side.
Do the Things AI Absolutely Cannot Do
Going one step beyond “worth the most,” the thesis establishes this as an active choice rather than a passive defense: not to fight AI over what it can do, but to go straight for “the things AI absolutely cannot do.” The ground is the divide between the phenomenal layer and the being layer — AI operates only within the phenomenal layer, whereas a human being can reach beyond the phenomenal. This follows directly from Form and Nature: The Phenomenal Layer Can Never Explain the Being Layer: all the ingenuity of the phenomenal layer can never explain being itself, and however powerful AI grows, it cannot take that step out of the phenomenal layer.
This choice carries the color of a calling, and its motive is not to prove anything outwardly: “We are precisely here to do the things AI absolutely cannot do — this is what we are here to do. I am not doing it in order to prove anything, because ever since I was a child there has been a voice within me. I am far too proud for that.” The key here lies in the source of the motive — one does this because of an inner voice, not out of a contest with anything outside. From this follows a criterion for selection: do “the things that will rise along with AI, not the things that chase after AI’s technology.” The former takes AI as a base that amplifies oneself — the more powerful AI becomes, the more one benefits; the latter is a chase after the technology itself, doomed to be left behind by each new iteration. This orientation is the obverse of Train the AI Mindset, Not Coding: A Systematized Solution Beats Any Tool and Rises with the Infrastructure.
Awakening Is AI’s Eternal Boundary
Among all “the things AI cannot do,” one definite, impassable boundary is named: awakening. “AI’s boundary. The thing AI can never do is awakening — and I must do it.” Awakening is AI’s absolute boundary for the same reason that grounds this whole thesis at the being layer: awakening is not the piling up of more information, but the very act of seeing through appearance itself and withdrawing from the phenomenal layer — and AI is trapped inside the phenomenal layer, with no vantage point from which anything could be “seen through.” This is the core of AI Cannot Awaken: Meditation Is the One Thing AI Can Never Replace, and it shares its mechanism with the “to see through is to be free” of To Know the Illusion Is to Leave It: To See Through Is to Be Free.
To set up awakening as AI’s boundary is, at the same time, to answer the question of where a person should direct their effort: since this is a place AI can never reach, it is the direction of investment in which a person will least depreciate. It converges with the “rise along with AI” criterion — the deeper one’s awakening, the more one can command every ability that AI can amplify, while awakening itself is never touched by any technological iteration.
The True Moat Is Cost
Why can AI not even do the thing of “becoming a certain person”? The answer is located in cost, not in perspective or in style. Even if AI can perfectly imitate a person’s style, it still cannot truly “become” that person — “because it has no capacity to pay a price. It will not die, will not lose, will not regret.” This is the sharpest edge of the thesis: people commonly suppose the moat is a unique perspective or aesthetic, but a perspective can be learned and a style can be replicated; what truly cannot be copied is the irreversible price congealed behind a work.
Death, loss, regret — these are the limits of being human, and they are precisely the capacities AI can never acquire. A being with no risk of loss makes choices that carry no weight; only a person who can lose gives their choices and their works any real heft. From this, cost is not a cost to be overcome but the very source of value — the same logic as Suffering Is the Tempering of the Soul: the tempering holds precisely because what it demands is something irretrievable. The moat that cost constructs is also where The Great Inversion of Value: AI Levels Cleverness, and Causality, Kindness, Wisdom, Faith, and Philosophy Become Worth the Most comes to rest at the level of the individual.
Value Lies in “Becoming”
Cost explains why AI cannot become a person who “already is”; and the final layer of the thesis goes further, pointing out that a person’s deepest irreplaceability lies not in what he “is” but in what he is “becoming.” The formulation: “Your value lies not in what you ‘are’ but in what you are ‘becoming.’ This is a process forever in the unfolding. Machines will grow ever smarter, but they are forever chasing a moving target — and that target is you, if you are still moving.”
This layer transforms irreplaceability from a static attribute into a dynamic relation. Any “already-fixed” ability will, sooner or later, be caught and surpassed by AI; but a person still growing, still unfolding, is a moving target — what the machine catches is always his old version, while he has already moved on to a new position. There is a precondition hidden here, added in the same breath: “if you are still moving.” The moat is not innate; the person who stops growing is instantly caught up with. This echoes Depth of Thought Cannot Be Replaced: AI Filters Out the Shallow Influencers, and the Darker the Sky the Brighter the Stars — what AI filters out is exactly those who have stopped moving and merely repeat themselves at the phenomenal layer; and it connects with What Is Scarce Is the Capacity to Carry Meaning: Narrowing Is a Bargain, and the Non-Standardizable Is Scarcer Still: the non-standardizable is so precisely because it is still in the making.
Here the several layers of the thesis draw together into a whole: what AI can do will be leveled, so what is worth the most is what AI cannot do; AI cannot do it because it is trapped in the phenomenal layer, has no capacity to pay a price, and cannot awaken; and a person’s ultimate moat is to live himself into one who is still advancing through cost, forever one step ahead of the machine — a “becoming.”
Sources
- Manuscript — “Is there anything AI cannot do? My friends — those are the things worth the most. And yet AI can still help you do the things AI cannot do.”
- Manuscript — another record of the same formulation
- Manuscript — “We are precisely here to do the things AI absolutely cannot do… I am not doing it in order to prove anything, because ever since I was a child there has been a voice within me. I am far too proud for that.”
- Manuscript — “The thing AI can never do is awakening — I must do it. And, the things that will rise along with AI, not the things that chase after AI’s technology.”
- Manuscript — “It has no capacity to pay a price. It will not die, will not lose, will not regret”; “Your value lies not in what you ‘are’ but in what you are ‘becoming’… it is forever chasing a moving target.”
See also
- AI Is a Machine of Probabilistic Permutation: It Only Fills In the Phenomenal Layer and Is Never Itself Inspired
- AI Cannot Awaken: Meditation Is the One Thing AI Can Never Replace
- Depth of Thought Cannot Be Replaced: AI Filters Out the Shallow Influencers, and the Darker the Sky the Brighter the Stars
- The Great Inversion of Value: AI Levels Cleverness, and Causality, Kindness, Wisdom, Faith, and Philosophy Become Worth the Most
- What Is Scarce Is the Capacity to Carry Meaning: Narrowing Is a Bargain, and the Non-Standardizable Is Scarcer Still