LETTER 063·The Independent Artist

Why algorithms cannot simply find the best music

Algorithms cannot simply find the best music because they optimize engagement signals, not quality — discovery favors what captures attention, not what deserves it.

August 30, 2026

Algorithms cannot simply find the best music because they do not measure quality; they measure engagement signals, which correlate with quality only loosely and often inversely. Discovery is optimized for capturability, not merit.

Quality is not the signal

Algorithms optimize what they can measure — clicks, completion, repeats, shares. None of these are measures of artistic quality; they are measures of behavioral response. A work can be shallow and capture attention, or profound and not — and the algorithm will favor the first.

The wrong optimization

This is not a flaw to be fixed but a structural feature: a system optimizing engagement cannot, by design, optimize merit, because merit is not the signal it reads. The best music is not a quantity an engagement algorithm can locate.

In essence

Algorithms cannot find the best music because they optimize engagement signals, not quality, and the two correlate only loosely. Discovery favors what captures attention, not what deserves it, because capturability is what the system measures.

Related questions

Can algorithms be improved? They can optimize different signals, but no engagement signal is equivalent to quality.

Is human curation better? Often, for quality, but it does not scale — which is why algorithms, despite their limits, dominate.

ENDERUM

© 2026 Enderum