Why execution can be delegated, but judgment cannot.
Artificial intelligence challenges traditional notions of authorship and judgment. This foundational essay explores why execution can be delegated — but genuine intellectual responsibility cannot.
During the Renaissance, artistic creation rarely conformed to the modern myth of solitary genius. Master painters worked in workshops. They conceived the composition, drafted initial sketches, set proportions, and ensured coherence. Apprentices executed large sections of the work — preparing surfaces, applying pigment, and completing secondary figures.
What mattered was not who held the brush, but who held the vision.
Authorship did not depend on the number of brushstrokes personally applied.
It depended on intellectual architecture.
Execution could be delegated.
Conception could not.
What distinguished a Leonardo or a Raphael was not manual exclusivity but intellectual architecture: the capacity to conceive a form, organise complexity, and impose unity upon multiplicity.
The analogy with artificial intelligence is not superficial; it is structural.
Today, AI systems generate text, images, code and analysis at extraordinary speed. They extend productive capacity as much as apprentices once did. They accelerate execution, expand output and assist refinement.
But they do not originate the vision.
They do not determine which questions are worth asking, which distinctions matter, or which direction inquiry should take. They operate within parameters set by human intention. They process patterns. They do not inhabit meaning. There lies the limit.
The architecture remains ours — or it dissolves.
This distinction becomes especially urgent in education.
A recurring anxiety has emerged: when students use AI tools for academic work, does authorship disappear? Has originality been compromised?
The concern is understandable, but it is also incomplete.
Plagiarism, in its classical sense, concerns the misappropriation of identifiable words or ideas. Generative systems complicate this definition because they produce novel formulations rather than copying from traceable sources. From a purely textual perspective, the output may be technically original.
But education has never been merely about textual originality. It has been about intellectual formation.
Authorship, in its deeper sense, lies not in phrasing but in ownership of questions, responsibility for arguments, and the capacity to justify conclusions. It resides in the tacit dimension of understanding — the integration of evidence, context and judgement that cannot be fully reduced to explicit procedure.
If a learner defines the problem, evaluates evidence critically, and assumes responsibility for interpretation, AI may serve as an instrument — like the apprentice in a Renaissance workshop. If the machine performs the thinking and the human merely submits the output, authorship becomes hollow, even if the wording is new.
The decisive distinction is therefore not between “AI” and “no AI.”
It is between assistance and substitution.
Used reflectively, AI can clarify structure, expose weaknesses in reasoning and stimulate further inquiry. It can serve as cognitive scaffolding. Used uncritically, it risks encouraging passivity — replacing the very processes academic work is meant to cultivate: doubt, revision, discernment and judgement.
The anxiety surrounding AI often masks a deeper discomfort. We fear learners will cease to think for themselves. Yet history suggests that tools do not determine outcomes; frameworks do. Calculators did not abolish mathematics. Digital archives did not abolish research. Each technological shift compelled a clearer articulation of what intellectual responsibility entails.
Artificial intelligence now performs a similar function. It serves as a stress test for educational design. If a task can be fully outsourced to a generative system without intellectual loss, perhaps it never required genuine judgment in the first place.
This is the uncomfortable implication.
There was no golden age of critical thinking suddenly destroyed by algorithms. What exists is a moment that exposes fragilities. AI does not create the weakness. It reveals it.
The central ethical question is therefore no longer whether a text is original. It is whether the thinking behind it is genuinely human — reflective, accountable and internally owned. Judgement is not merely a cognitive operation; it is a form of responsibility.
In the Renaissance workshop, greatness lay not in the brush but in the mind that guided it. In the age of artificial intelligence, the same truth holds.
Technology expands what we can calculate. Education must strengthen our capacity to judge.
Execution can be delegated. Judgment cannot.
In the age of intelligent machines, preserving human judgment is not resistance to technology. It is the condition for using it well.
Developing this argument
- AI as a Stress Test for Critical Thinking →
- AI, Plagiarism, and Intellectual Authorship in the IB Diploma Programme →