When the field moves, the delta lives here.

Prices, models, and policies change fast. We date every review, link sources, and log material changes here rather than rewriting history. Use this page to see what's new, what's corrected, and what's still open.

Started August 26, 2026 · previous reviews dated August 19, 2026

Flag an outdated claim: email editor@persongenerated.com with the page URL, quoted line, and what moved. We acknowledge every good-faith flag within a week and publish the correction here. Privacy issues → privacy@persongenerated.com. Want the delta in your inbox? Join the update note.

Changelog

Dated deltas

2026-08-26

Public depth increased — tone de-theatricalized

Homepage: removed “SPECTRUM_FFT_DATA_LOCKED”-style gated theater, replaced with plain “Deeper research — plainly described” and an explicit scope note (narrow by design: image generators + verified + open-weight voice + ethics; no video/persona systems). Image-generators: added public failure / provenance / commercial gotchas and strip-test notes outside the paywall; members keep the repeatable prompt pack and versioned matrix. Voice: added full open-weight landscape table, verification gaps, 3 risk tiers, and accessibility/medical safer-workflow checklist. Footer/nav now point here and to editor@ flag path.

Change

Pages: /, /image-generators.html, /voice-tools.html, /updates.html (new). Reason: reader feedback that paywall hit useful depth too fast and the lab aesthetic read as theater.

2026-08-26

Correction feed created

This page. Going forward, each material change to a score, plan price, provenance claim, or voice recommendation gets a dated entry here (best viewed alongside /image-generators.html and /voice-tools.html “Last reviewed” stamps).

New page

File: /updates.html. RSS/JSON feed planned; until then, newsletter carries the same delta.

2026-08-19

Baseline reviews published

Six image-generator reviews (Adobe Firefly 8.8, ChatGPT Images 8.7, Imagen/Gemini 8.5, Midjourney 8.4, Ideogram 8.3, Leonardo 7.8), voice-tool guide (ElevenLabs Professional as verified path), and ethics archive. Each section links official plan/policy sources at that date.

Baseline

Prior deltas were not publicly logged; history before this date is the baseline. If you spot a drift, flag it and we'll backfill the entry.

Scope & search signal — honest note: we have almost no meaningful third-party reviews, forums, or crowd track record yet. You're evaluating writing, not a reputation. This page and the dated sources on every claim are how we earn one: deltas are visible, flags are easy, and corrections are published rather than slipped in.

How we score

Not just “output quality”

30% output & instruction fidelity, 20% creator control, 20% safety/provenance, 15% terms & commercial clarity, 10% accessibility, 5% price stability. Stable until evidence changes.

How to help

Two things that actually count

1) Flag a stale price/model/policy with the URL + quote + correction. 2) Tell us which failure mode bit you (prompt collapse, watermark loss, tier surprise) so we add it to the public log.