Sustainable Open-Source AI Requires Tracking the Cumulative Footprint of Derivatives
Shaina Raza, Iuliia Zarubiieva, Ahmed Y. Radwan, Nate Lesperance, Deval Pandya, Sedef Akinli Kocak, Graham W. Taylor · 2026
Read Sustainable Open-Source AI Requires Tracking the Cumulative Footprint of Derivatives by Shaina Raza, Iuliia Zarubiieva, Ahmed Y. Radwan, Nate Lesperance, Deval Pandya, Sedef Akinli Kocak, Graham W. Taylor (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-21632.txt.
Description
Shaina Raza and six co-authors argue that open-source AI must track the full footprint of model families, not just single models. Hubs now host millions of artifacts. Each base model can spawn fine-tunes, adapters, compressed copies, merges, and forks by the thousands.
Judging sustainability by the cost of one training run misses that sprawl. The authors take the position that the cumulative footprint of all derivatives is the honest ledger. Counting the family, not the founder, changes which practices look green. It also pushes hubs and builders to report totals that match the real energy story.
Catalog record
| CREATOR | Shaina Raza, Iuliia Zarubiieva, Ahmed Y. Radwan, Nate Lesperance, Deval Pandya, Sedef Akinli Kocak, Graham W. Taylor |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-21632