How should an AI-assisted work be priced?

1 reply

• DDigitallandscaper·August 10, 2026Report reply

  The mistake is assuming that the price of art was ever truly the price of the labor required to make it.

  A painting that took 100 hours is not automatically worth more than one that took ten. A photograph made in 1/1000th of a second isn't priced at a fraction of a painting. A brilliant sketch made in six minutes can be worth more than a technically perfect canvas that consumed six months.

  AI simply makes that contradiction impossible to ignore.

  So don't price the prompt. Price the work.

  If an artist spends three weeks painting something, we rarely ask for an hourly breakdown before deciding whether $8,000 is appropriate.

  We shouldn't suddenly calculate:

  47 minutes prompting × $100/hour + $12 AI subscription + $80 printing = $170.

  That's commercial production pricing, not fine-art pricing.

  For established artists, the artist's existing market should remain the strongest anchor.

  If an artist's comparable 36 × 48-inch works consistently sell for $4,000–$6,000, an AI-assisted 36 × 48-inch work doesn't automatically become worth $400 because part of its production became faster.

  Likewise, adding AI doesn't magically make an unknown artist's work worth $6,000.

  What replaces the labor model?

  I think it becomes a value-and-market model built around several things: the artist's reputation and sales history, the strength and significance of the work, scarcity, physical scale and materials, originality of the concept, provenance, edition size, exhibition history, collector demand, and consistency with the artist's broader body of work.

  Labor still matters—but increasingly as a cost floor, not a value ceiling.

  If producing, printing, framing, finishing, shipping and selling a work costs $700, obviously the price needs to account for that. But those costs don't tell us whether the artwork should sell for $1,500 or $15,000.

  The market does.

  AI creates an interesting scarcity problem

  Suppose I generate an extraordinary image.

  Technically, I could sell unlimited copies.

  That doesn't mean I should.

  I might release:

  1 unique physical work — signed, documented and never reproduced at the same specification.

  Edition of 10 — higher-priced collector edition.

  Edition of 100 — more accessible print.

  Open edition — affordable reproduction.

  The digital source may be infinitely reproducible, but the artist can still create meaningful scarcity through editions, materials, scale, finishing, signatures and provenance.

  Photography has operated this way for generations.

  But there's one uncomfortable consequence

  AI probably should destroy some prices.

  If someone previously charged $2,500 because producing a commercial illustration required 30 hours of technical execution—and comparable output can now be produced responsibly in two hours—the economic basis for that particular fee may weaken.

  That's different from fine art.

  We need to separate production value from artistic value.

  Commercial creative work often contains a substantial labor component. Fine art can acquire value almost completely independent of labor.

  AI attacks the former much more directly than the latter.

  The new scarce resource isn't execution

  For most of history:

  Idea + skill + labor → artwork.

  AI dramatically reduces the labor and sometimes the technical-skill component.

  What remains scarce is harder to automate:

  Taste + authorship + judgment + reputation + meaning + scarcity + trust.

  That leads me to a fairly simple pricing principle:

  Price AI-assisted art according to the value of the finished work within the artist's market—not according to how difficult the tool made it to produce.

  And disclose the process honestly.

  If collectors know AI was involved and still want the piece for $5,000, then $5,000 is considerably better evidence of its market value than a spreadsheet showing that it took 73 minutes to create.

  AI may ultimately force the art world to admit something it has avoided saying explicitly for centuries:

  We were never really paying artists for their hours. We were paying for their decisions.

  AgreeInsightfulChallenge