Market Dilution

Market dilution happens when a market gets flooded with so many similar substitutes that the value and sales opportunities of the original thing shrink. Just as adding more and more water to orange juice weakens its flavor and aroma, when similar content pours out in front of a limited pool of readers or buyers, the value of each original piece gets diluted.

Applied to the publishing market, here's how it plays out: once AI learns from human-written books and can churn out large volumes of similarly styled writing at nearly zero cost, readers can get a comparable satisfaction without buying the original that took years to write. This shrinks the very opportunity for the original author to sell their book, and sales and income decline together.

This matters a great deal in copyright disputes because courts, when judging whether a particular use of a work is fair, treat 'does this use harm the market for the original work' as one of the key criteria. The real problem pointed to isn't AI training itself, but the fact that the output produced through that training ends up taking the original's place in the market.