RETROSPECTIVE RECORD · PREPARED 16 SEPTEMBER 2026The archive · 100 retrospective records ↗
← The archive

Eight songs on one album named four different AI tools

The artist's own site credits Amper, Watson Beat, AIVA and NSynth track by track for the 2018 album I AM AI.

Historical event
September 1, 2018
Site publication
September 18, 2026
Visual for this record: Eight songs on one album named four different AI tools
Visual published by livenewschat.eu, shown for identification of the record. Credit: livenewschat.eu · source page ↗ Rights: owner-review-pending. Source

What happened

In September 2018 the Los Angeles-based artist Taryn Southern released I AM AI, an eight-track pop album she describes on her own site as composed and produced with several separate AI systems rather than one. The page lists, song by song, which tool supplied which part: Amper for "Break Free", "Life Support" and "Time"; IBM's Watson Beat combined with Amper for "New World" and "Alive"; Watson Beat alone for "Genesis"; AIVA for "Lovesick"; and Amper alongside Google's NSynth for "Voices". Her about page calls it "the world's first solo pop album composed with artificial intelligence".

What the documents say

The album page describes a consistent division of labour across all eight tracks: the AI software composed notation and instrumentation, while Southern wrote lyrics and vocal melodies and handled arrangement, and producer Ethan Carlson is credited with "vocal production, mixing and mastering". That split matters because it names what the AI tools did and did not do; none is credited with lyrics, vocal performance or final mix decisions. Her about page adds a reception claim in her own words, that "Break Free" reached "the Top 100 US radio charts (Mediabase)" and drew coverage from Wired, Forbes and the BBC. Neither page discloses what material trained Amper, Watson Beat or AIVA; the credits establish which tool touched which song, not what each tool learned from.

Why it matters for makers

Crediting four tools by name and by track is the useful practice here, and it is rarer than it should be. A listener can see, from the artist's own page, exactly which layer came from which generator, rather than a single blanket "made with AI" label that obscures whether a program wrote a melody or merely rendered one the artist specified. For a composer weighing several AI tools on one project, mixing tools by function, one for notation, one for instrumentation, one for a song's mood, is itself a compositional decision distinct from what any single tool produces.

What to check before you use it

A maker crediting multiple AI tools on one release should record, per track, which tool contributed which layer, as Southern's page does, so a later rights or royalty question has a clear map rather than a memory. It is worth checking each tool's own terms for commercial output rights before release, since a tool used for one song's instrumentation may license differently from one used for another's melody. This is an editorial checklist; neither source states these are legal requirements.

I AM AI is a useful record precisely because it resists treating "AI music" as one undifferentiated thing. The artist's own page distinguishes four tools by name and task, a level of specificity a maker can learn from even where the underlying training data stays undisclosed.

Sources & reading trail

Song-by-song listing of which AI tool (Amper, Watson Beat, AIVA, NSynth) was used, and description of the production process.

Source published: Not established · Retrieved: 16 September 2026

Artist's own description of I AM AI as AI-composed and her account of the single's chart performance and press coverage.

Source published: Not established · Retrieved: 16 September 2026

Papers, terms and official documents establish the record; the maker reading and the checks are Signal to Song editorial analysis. This retrospective draft does not imply the site published on the event date.

Continue reading

Sources & reading trail

The documents above establish the record. The reading and the questions are this publication’s editorial analysis, written after the fact.

Published September 18, 2026, not on the date of the event described.