Purpose of this document
This document describes how HPR editors proactively find candidates for the registry. It answers the question: “Why did we include this author and not that one?” The honest answer is that our scouting starts with sources that are statistically likely to yield human voices, but finding a candidate through a scouting source never substitutes for independent verification. Every candidate must satisfy both evidence pillars regardless of how they were found.
This document is published openly because our scouting methodology is part of our editorial transparency commitment.
The guiding principle of scouting
We are not looking for prestigious authors or widely read publications. We are looking for human voices. Prestige and readership are irrelevant to the registry’s purpose. A blogger with 200 readers who has been publishing daily for 12 years in a distinctive personal voice is a better candidate than a journalist at a major publication who joined their outlet in 2023 with no prior archive.
What editors look for when scouting: A candidate is worth pursuing when we find: a substantial retrievable archive of prior work predating 2022; stylistic consistency across that archive; personal and biographical specificity that accumulates into a portrait of a specific person; and no red flags suggesting AI involvement in the prior work.
Scouting source 1: Journalism awards databases
Sources used: Pulitzer Prize (pulitzer.org), George Polk Awards, IRE Awards, Online Journalism Awards, SPJ Sigma Delta Chi Awards, ASME National Magazine Awards, Overseas Press Club Awards.
How we use them: Award databases identify journalists who received independent third-party recognition for specific works. We use them as starting points — a named journalist with a Pulitzer citation has a documented work and a verifiable identity. That is the beginning of our review, not the end of it.
Additional verification always required: Being award-winning does not satisfy either evidence pillar. After identifying a candidate through an award database, editors must: verify the author’s prior archive independently; assess stylistic consistency; and confirm that the awarded work itself reads as consistent with that prior human voice. Award-winning journalists have been found to use AI; the award is context, not evidence.
Known limitations: Awards skew heavily toward US and English-language journalism, toward staff journalists at large outlets, and toward certain beats (politics, investigation). They underrepresent independent journalists, international voices, and specialist or niche writing. Editors must compensate for this bias actively.
Scouting source 2: Long-running independent blogs
How we find candidates: The Wayback Machine CDX API (web.archive.org/cdx) allows batch retrieval of archive records for any URL, providing timestamp evidence of when a blog was actively publishing. Third-party blog directories, Hacker News “Ask HN” threads about recommended blogs, and blogrolls from already-listed bloggers provide candidates.
What makes a blog a strong candidate: A blog with hundreds of archived snapshots spanning 2010–2021 has a decade of pre-AI content that establishes a baseline human voice. When we find such a blog, we read across the archive — early posts, posts from the middle period, and recent work — looking for continuity of voice.
Known limitations: Many long-running blogs have become inactive; some have changed authors without announcement; some have published inconsistently. An old domain does not guarantee an active human author today.
Scouting source 3: Newsletters (Substack, Ghost, independent)
How we find candidates: Substack’s discovery pages, the third-party directory Sidestack.io, Ghost’s featured publications list, and recommendations from existing listed authors. We focus on newsletters whose authors have a verifiable publishing history predating their newsletter — typically journalists, academics, or bloggers who moved to newsletter format but whose prior work is retrievable.
What makes a newsletter a strong candidate: The author has a verifiable prior publishing history (journalism, blogging, academic writing) going back at least five years before the newsletter launched, and the newsletter voice is consistent with that prior work. An author who appeared on Substack in 2022 with no prior verifiable history is a poor candidate regardless of newsletter quality.
Known limitations: Substack discovery surfaces popular newsletters, not necessarily the most credibly human ones. Large subscriber counts are irrelevant to our assessment.
Scouting source 4: Literary and essay publications
Sources used: Longreads, Aeon, Nautilus, Guernica, The Rumpus, n+1, Electric Literature, Granta, The Believer, The Paris Review, Lapham’s Quarterly, The Yale Review, Salmagundi, and similar venues.
How we use them: These publications commission and publish personal essays, reported essays, and literary criticism where authorial voice is the primary value. Authors who appear regularly in these venues over multiple years, whose work is stylistically consistent, and who have a retrievable archive are strong candidates.
Additional verification required: We verify each author’s prior archive independently. We do not assume that publication in a literary journal implies human authorship — these venues have also faced AI submission floods and have varying levels of editorial scrutiny.
Scouting source 5: Academic publications
How we find candidates: Scholar profiles on Google Scholar, ResearchGate, and institutional pages; references in academic work already listed in the registry; and humanities and social science journals relevant to the registry’s topical scope (media studies, journalism studies, science and technology studies, digital humanities).
What makes an academic a strong candidate: A scholar with a Google Scholar profile showing publications spanning 10+ years, with a consistent research focus and writing style, and whose work includes the kind of personal voice and engaged argument that characterises human academic writing.
Known limitations: Academic writing is now heavily compromised by AI use. The existence of a publication record does not establish human authorship of any specific paper. Editors must read the actual work, not just the citation record, and assess stylistic consistency carefully.
Scouting source 6: Social media accounts
How we find candidates: We look for specific postsm not just accounts, by writers who already appear in the registry or who have strong independent verification through other scouting sources. A Substack Notes post by an author with a 10-year blog archive is a strong candidate. A LinkedIn article by a journalist with a verifiable 15-year byline history is a strong candidate.
What editors assess: The specific post must be substantive original content (not a reshare or brief comment). The post must be consistent with the author’s prior voice. The author’s verifiable history must satisfy Pillar 1 requirements.
Known limitations: Social media is the easiest environment for AI-generated content. We are highly selective about social media entries and require stronger corroborating evidence than for other content types.
Scouting source 7: Community nominations
Community nominations are reviewed against identical standards to scouted entries. Nominators are asked to provide: the specific entry, the prior archive evidence, and their assessment of which evidence pillar applies. Strong nominations include specific URLs demonstrating the prior archive and a brief note on why the voice is consistent.
Editors do not give community nominations preferential or disfavoured treatment relative to scouted entries. The evidence is assessed on its merits.
Diversity tracking
HPR editors maintain a running record of the geographic, topical, and institutional distribution of entries. At each quarterly editorial review, the team assesses whether scouting is over-representing certain categories, US journalism, technology criticism, major outlets, and actively redirects scouting effort toward underrepresented areas. Target areas for active expansion include: international English-language journalism (UK, Australia, Canada, Ireland, South Asia, Africa); independent voices in fields underrepresented in the current registry; women writers and writers from underrepresented communities; and non-technology topical areas.
Recording scouting decisions
Every candidate reviewed and declined ,not just accepted, is recorded in the editorial log with: the candidate’s name or URL, the scouting source, the reason for decline, and the reviewing editor. This log is internal but available to the editorial board for audit. Patterns of decline are reviewed quarterly to identify whether the evidence standards are being applied consistently.