AI / AUTHORSHIP / RESPONSIBILITY
The tool can propose.
The human must answer.
These projects have been developed through AI-assisted experimentation. Treating the collaboration clearly is part of the craft, not an apology for it.
EDITED JUDGMENT ↗
OVER AUTOMATED PRESTIGE
What AI can help with
Drafting code, exploring layouts, accelerating prototypes, proposing interaction patterns, and generating alternatives for text or structure. A useful draft can be an invitation to learn. It is not, by itself, proof that every feature works as intended or every statement is accurate.
What remains a human responsibility
Deciding which question deserves a page; establishing what the page is permitted to claim; checking sources and links; editing the argument; respecting people depicted or implied by the work; and deciding whether an experience should be published. Accessibility should be tested rather than inferred from attractive markup.
Where the evidence stops
MADE cannot reconstruct an exact tool-by-tool authorship log for every existing exhibit. It therefore makes no percentage claim about human versus machine contribution, and it does not present AI-assisted output as hand-coded from scratch. The content here was newly edited for this exhibition, and links were reviewed at the level described in the notes below.
The choice that matters
Neither “AI-made” nor “human-made” is sufficient as an evaluative category. A better question is: can another person inspect the artifact, understand its limits, learn something from it, and identify who is accountable for errors? In the best version of this practice, speed creates room for more judgment—not an excuse to skip it.