Why Authentic Voices Matter in Podcasting and the Open Web  🔗 **Source:** [Why Authentic Voices Matter in Podcasting and the Open Web](https://openchannels.fm/why-authentic-voices-matter-in-podcasting-and-the-open-web/) > [!note]+ Document Notes > tags: [podcasting, ai-disclosure, authenticity-claims, open-web, content-provenance, creator-economy] > author: Bob (host, Open Signal) > source: Open Channels FM > reference: https://openchannels.fm/why-authentic-voices-matter-in-podcasting-and-the-open-web/ > date created: 2026-09-11 > > > [!info] Generated analysis — not the author's words. Sections below are machine-written > > from the captured source. Only text inside `> [!quote]` blocks is verbatim from the author. > > Do not quote or attribute any other line to the named author. > > > [!note] Attribution and capture > > The page carries an **unfilled template placeholder — `[Guest Name]` — where the byline belongs**, > > so the site does not name the author. The host self-identifies in the audio only as "Bob," which is > > what this note uses; no surname is asserted. Reader captured the full 525-word transcript; the live > > page's additional ~800 words are player UI, share links and embed code, so nothing substantive was > > missed. Episode runtime 3:35. > > # 🏷️ Title: Why Authentic Voices Matter in Podcasting and the Open Web > <!-- deep-read:sections start --> ## 🧭 Table of Contents - [📝 Summary](#-summary) - [📖 Main Content](#-main-content) - [🗒️ Notes](#️-notes) - [📚 References](#-references) - [🗣️ Paraphrase](#️-paraphrase) - [📰 Newsletter](#-newsletter) - [✅ Action Items](#-action-items) - [🧪 Hypotheses](#-hypotheses) - [📊 Methodology](#-methodology) - [❓ Questions](#-questions) - [🔮 Future Research](#-future-research) - [💡 Implications](#-implications) - [🧠 Learning Styles Assessment](#-learning-styles-assessment) - [📊 Accessibility Assessment](#-accessibility-assessment) - [🌍 DEIB Assessment](#-deib-assessment) - [✅ Logical Validity](#-logical-validity) - [🛡️ DARVO Analysis](#-darvo-analysis) - [🎯 Conclusion](#-conclusion) - [🔄 Refactor Appendix](#-refactor-appendix--why-authentic-voices-matter-in-podcasting-and-the-open-web) ### 🔗 References - [🗒️ Notes — the distinction that carries the episode](#️-notes) - [✅ Logical Validity — an unverifiable claim, honestly made](#-logical-validity) - [🔄 Refactor Appendix — atoms and permanent candidates](#-refactor-appendix--why-authentic-voices-matter-in-podcasting-and-the-open-web) ## 📝 Summary - **What happened.** A 3:35 solo episode of Open Signal in which the host states that neither this show nor any podcast on the Open Channels FM network is AI-generated, while being explicit that he does use AI — for repurposing finished episodes into show notes, takeaways, and blog posts. - **Why it matters.** The useful content is one line drawn precisely: AI applied to material a human already made is a different act from AI producing the material. That distinction is more usable than a blanket "we don't use AI," and more honest than not mentioning it. - **Key stakeholders.** Podcast listeners deciding what to trust; independent creators competing against generated audio; anyone writing an AI-disclosure policy for their own publishing. - **Risks and tradeoffs.** Every claim here is self-attested and unverifiable. There are no external links, no evidence, and no mechanism offered by which a listener could confirm any of it — including the closing assertion that the script was not AI-written. - **Bottom line.** Take the distinction, not the assurance. The generation-versus-repurposing line is genuinely reusable; the episode's own truthfulness is something a listener can only choose to accept, and the episode does not pretend otherwise. ## 📖 Main Content **The claim.** The host builds on a prior episode about why the show is solo — *"Just me, you know, no panels, no hedging"* — and adds what he says he did not state outright before: that it is not only his perspective but genuinely him. Across eight years of the site's content and across the network's shows, he says the podcasts were never AI-generated: *"There isn't a script or an idea that started as a prompt."* **The distinction that makes it worth recording.** He does not claim AI abstinence. He says plainly *"I do use AI. I'll be straight about that,"* and specifies where: once an episode is finished, for show notes, key takeaways, questions asked, and sometimes a blog post pulled with a particular emphasis. Then the line: *"That is simply AI doing something useful with content a human already made. It's really not the same as AI making the content in the first place, or what a lot of people call slop."* That is a cleaner formulation than most AI-disclosure statements manage. It separates the question "was AI involved" — increasingly useless, since the answer is nearly always yes somewhere — from "did a human originate this," which is the question listeners actually care about. Stated as a general rule it would apply to writing, video, and code as readily as to audio. **The positioning, and its limits.** The episode's emotional centre is *"there's still a person behind the voice when you listen to any of our episodes"* and *"It's not a format, not a workflow, just me talking."* He explicitly disclaims superiority — not boasting, not knocking anyone doing it differently — which is the right register for a claim that could easily curdle into moralizing, and largely keeps it there. **The unverifiable close.** The final beat is a self-referential proof: *"this was not scripted by AI. I actually wrote this script down and did it myself."* This is the episode's structural problem in one sentence. An AI-generated script asserting it was not AI-generated would read identically. The claim is unfalsifiable from the artifact alone, and no provenance mechanism — signed content, C2PA-style metadata, a published process — is offered. That is not dishonesty; it is the current state of the field. It does mean the episode's value is the argument rather than the assurance. **What the page itself undercuts, mildly.** The published page carries an unfilled template placeholder, `[Guest Name]`, where the author byline should be. On a page arguing that a real person stands behind the work, the field naming that person was never filled in. It is a CMS oversight rather than a contradiction, and worth noting once without making more of it than it deserves. The `<title>` tag also differs from the page heading — "Why Authentic Voices Matter in a Changing Audio Content World" versus "…in Podcasting and the Open Web." **On "the open web" in the title.** The phrase appears in the headline and is never developed in the audio. Nothing in the 525 words addresses open protocols, syndication, RSS ownership, or platform independence. Whatever the open-web argument was meant to be, this episode does not make it. ### 🗒️ Notes - **Runtime 3:35**; 525 transcript words. A short, deliberately informal monologue, not an essay. - **The core distinction:** AI repurposing finished human work ≠ AI generating the work. Named uses are show notes, key takeaways, questions asked, and occasional derived blog posts. - **Scope of the claim** extends beyond the host to "my hosts on all the podcasts we have" — a claim about other people's practice, made on their behalf without their voices present. - **Eight years of site content**, and the host says he has been making some form of content since the late '80s. - **"Slop"** is used as the received term for generated content, without definition — assumed shared vocabulary. - **Explicit non-superiority disclaimer**, twice: not boasting, not knocking others. - **Zero external links** in the captured content. Verified against the markup, not inferred from stripped text — this is a genuine absence. - **No evidence offered** for any claim, and none of the standard provenance mechanisms mentioned. - **`[Guest Name]` placeholder** unfilled on the live page where the byline belongs. - **Title-tag/heading mismatch** between "a Changing Audio Content World" and "Podcasting and the Open Web." - **"The open web"** appears only in the title; the audio never addresses it. - **Reader's capture is complete** for analytical purposes — the live page's extra words are player chrome and embed markup. ## 📚 References - [The episode page](https://openchannels.fm/why-authentic-voices-matter-in-podcasting-and-the-open-web/) — audio, 3:35, with subscribe links for Apple Podcasts, CastBox, Overcast, PocketCasts, Spotify and RSS - The prior Open Signal episode on why the show is solo — referenced as "last Friday" but not linked from this page - Also in this vault and directly adjacent: the CCM-industry series' episode on AI-generated worship music (Jonathan Allen Wright, 2026-09-09), which treats the same generation question in a different market; and the Standard.site guide on publishing into an open protocol, which is what "the open web" in this title might have meant ## 🗣️ Paraphrase - The host follows up a prior episode about the show being solo by adding that it is not only his perspective but authentically him. - Across eight years and the whole network, he says no podcast content has been AI-generated and no idea began as a prompt. - He is explicit that he does use AI, specifically to repurpose finished episodes into show notes, takeaways, and blog posts. - He argues that AI working on material a human already made is categorically different from AI producing the material. - He disclaims any superiority over people who work differently, and says he raises it only because it is worth stating occasionally. - He has been making content since the late '80s and expects the forms to keep changing. - The constant, he says, is that a person is behind the voice in every episode. - He closes by stating the script was written by him and not by AI. ## 📰 Newsletter Three and a half minutes, one idea, and the idea is better than the running time suggests. The host of Open Signal spends the episode saying that nothing on his podcast network is AI-generated — but the interesting part is that he doesn't claim AI abstinence. He says outright: *"I do use AI. I'll be straight about that."* Once an episode is finished, it becomes show notes, key takeaways, sometimes a blog post. Then the line worth keeping: *"That is simply AI doing something useful with content a human already made. It's really not the same as AI making the content in the first place, or what a lot of people call slop."* That's a more useful disclosure than most. "Did you use AI?" is becoming a question with no informative answer, because nearly everyone did, somewhere. "Did a human originate this?" still means something. Splitting those cleanly is the contribution, and it transfers straight across to writing, video, and code. The catch is that none of it is checkable. There are no links, no evidence, and the closing flourish — *"this was not scripted by AI. I actually wrote this script down and did it myself"* — is exactly what an AI-written script claiming human authorship would also say. That isn't a knock on him so much as a statement about where provenance tooling currently is: signed content and content-credential metadata exist, and almost nobody uses them. One small irony worth a smile rather than a point: the published page carries an unfilled `[Guest Name]` placeholder where the byline goes. An episode about the real person behind the voice, on a page that never filled in the person's name. ## ✅ Action Items #### DevRel - [ ] Draft the generation-versus-repurposing distinction into a reusable AI-disclosure line for your own published work — context #things - Developer advocacy: this is a short, quotable framing for the disclosure conversation that keeps coming up. - Education: usable as a teaching example of a claim that is well-formed and unverifiable at the same time. - Developer experience: not applicable. - Developer marketing: authenticity as positioning is visible here in its honest form — worth contrasting with the performative version. - Community management: the non-superiority disclaimer is a good model for raising a practice standard without shaming. - Internal/external/cross-project: not applicable. #### OSPO - Policy & Governance: an AI-use policy that distinguishes origination from post-processing is more enforceable than a blanket ban. - License Compliance: not applicable. - Security Management: not applicable. - Software Inventory: not applicable. - Developer Support: not applicable. - Community Engagement: content-provenance claims will increasingly need mechanisms, not assertions. - Strategic Alignment: not applicable. #### Community - Maintainers / Contributors / Sponsors: not applicable. - Ecosystem: independent creators competing with generated audio is the market context, stated only implicitly. #### WordPress - Core / Core Performance / Core Program Team / Mobile / Playground / Photos / Polyglots / Openverse / Themes / Tide / CLI / Hosting / Plugins: not applicable. - Core AI: the origination/post-processing distinction is a candidate framing for any AI feature disclosure in core. - Community: not applicable. - Training: not applicable. - Marketing: not applicable. - Support: not applicable. - Design: the `[Guest Name]` placeholder is a template-fallback failure worth designing against. - Documentation: not applicable. - TV: WordPress.tv faces the same provenance question for submitted content. - Meta: not applicable. #### The WPOCC - Programs: not applicable. - Projects: a shared AI-disclosure norm for community media would be a small, achievable artifact. - Funding experiments: not applicable. #### FAIR - Technical independence: not applicable. - Policies and covenants: provenance and disclosure norms are covenant-shaped questions. - Productized governance artifacts: a disclosure template separating origination from post-processing. ## 🧪 Hypotheses - Disclosure statements that admit some AI use are more credible to audiences than blanket denials, because they signal a considered policy rather than a slogan. - The question "was AI used" will lose diagnostic value faster than "was this human-originated," as post-processing becomes universal. - Unverifiable authenticity claims accumulate until a provenance mechanism makes them checkable, at which point unverified claims become a negative signal. - Creators who have worked through multiple format shifts frame AI as one more shift, while newer creators frame it as a rupture. - Publishing platforms that ship template placeholders into production do so at a rate correlated with publishing cadence rather than with care. ### 📊 Methodology - **For the credibility hypothesis:** present listeners with matched disclosure statements — blanket denial versus qualified admission — and measure trust ratings. Straightforward survey work; nobody appears to have run it for audio. - **For the diagnostic-value hypothesis:** track how AI-disclosure questions are phrased in podcast directories and platform policies over time. Publicly observable. - **For the provenance hypothesis:** monitor C2PA or equivalent adoption in audio, then compare audience response to signed versus merely asserted content. Currently blocked by near-zero adoption. - **For the tenure hypothesis:** code creator framing of AI against years active. Requires a defined corpus; the framing is usually explicit enough to code reliably. - **Limits.** This is one 525-word monologue with no external evidence. It supports claims about *how* a creator frames the issue and nothing about the underlying facts, which are unverifiable from the artifact. ### ❓ Questions - What would let a listener verify any of this, and does the host consider that a problem worth solving? - Do the network's other hosts endorse the claim made on their behalf? - Where exactly is the line — is an AI-suggested episode topic origination or assistance? - What was the "open web" argument the title promises and the audio never makes? - Does repurposing with AI change the derived artifact enough that show notes should carry their own disclosure? - Would the host adopt a provenance mechanism if one were practical, or does he consider the stated commitment sufficient? ### 🔮 Future Research - Whether qualified AI disclosures outperform blanket denials on audience trust. - Adoption trajectory for content-provenance standards in independent podcasting. - How disclosure norms diverge between networks and solo creators. - Whether "slop" stabilizes as a defined term or remains a boundary-marking epithet. ### 💡 Implications - **Practice.** If you publish anything, the generation/repurposing split is the disclosure line worth adopting — it is specific, defensible, and survives the fact that you almost certainly use AI somewhere. - **Policy.** Blanket "no AI" commitments are becoming unenforceable and therefore not credible. Scoped ones are both. - **Security.** Not applicable, though provenance is the same class of problem as supply-chain attestation. - **Education.** A tidy case study in a claim that is well-constructed and structurally unverifiable — useful for teaching the difference between a good argument and established fact. - **Funding models.** Not applicable. - **For Courtney's work specifically.** Two connections. **First:** she publishes and podcasts in a community where this question is live, and the origination/post-processing distinction is directly adoptable for her own disclosure — particularly given how much of today's pipeline involves AI-assisted processing of sources she selected and read. This note's own provenance callout is the same move applied to itself. **Second:** the title promises "and the Open Web" and never delivers it, which is a gap the Standard.site guide processed earlier today actually fills — content portability, protocol-level ownership, and publishing that does not depend on a platform. Whoever writes the real version of this argument should read that one first. ### 🧠 Learning Styles Assessment - Reading level estimate: Flesch–Kincaid grade 7–8. Conversational spoken register with filler ("you know," "well"), short sentences, no jargon beyond "slop." #### 👀 Visual learners - Present: nothing in the transcript; the page has a player and episode art. - Improve: not applicable to an audio format, though a short written summary on-page would help. #### 🔊 Aural learners - Present: this is the native format — 3:35 of direct address. - Improve: none needed. #### 🗣️ Verbal learners - Present: strong — one clearly stated distinction, repeated and illustrated. - Improve: "slop" is used without definition. #### 🧩 Kinesthetic learners - Present: nothing to do. - Improve: no call to action beyond listening. #### 🧠 Logical learners - Present: the origination/post-processing distinction is a clean conceptual split. - Improve: no evidence, and the closing self-verification is circular. #### 🧑‍🤝‍🧑 Social learners - Present: direct second-person address throughout; acknowledges listeners will disagree. - Improve: no comment or response surface on the page. #### 🧘‍♀️ Solitary learners - Present: short and self-contained. - Improve: none needed. #### 🌿 Naturalistic learners - Present: none. - Improve: not applicable. ### 📊 Accessibility Assessment #### Screen reader semantics - Present: **a full text transcript is published on the page**, which is the single most important accessibility feature an audio publisher can provide, and it is genuinely good practice. - Improve: the `[Guest Name]` placeholder will be read aloud verbatim as page metadata, which is confusing rather than merely untidy. #### Keyboard navigation - Present: a custom podcast player with play/pause and speed controls. - Improve: custom audio players frequently have unlabelled controls and no keyboard scrub; not determinable from the capture and worth testing directly. #### Contrast and readability - Present: not assessable from the captured markup. - Improve: would require the live page. #### Text resizing and reflow - Present: transcript prose reflows cleanly. - Improve: none identified. #### Media and captions - Present: transcript provided — the correct equivalent for audio-only content, satisfying WCAG 1.2.1 in substance. - Improve: the transcript is not speaker-labelled, which matters little for a solo episode but would for interviews; no chapter markers. #### Forms and interactions - Present: subscribe and share controls. - Improve: share/embed markup appears inline in the extracted text, suggesting it may not be properly hidden from assistive technology. #### Standards reference - Transcript provision addresses 1.2.1 Audio-only well. The unfilled placeholder touches 1.3.1 Info and Relationships in spirit, and the custom player warrants a 2.1.1 keyboard check on the live page. ### 🌍 DEIB Assessment #### Cultural breadth - Present: none specific; the framing is generic to English-language independent podcasting. - Improve: no consideration of creators for whom AI assistance lowers a language or disability barrier — a significant omission in an argument about authenticity. #### Inclusive language - Present: notably careful. The non-superiority disclaimer appears twice, explicitly declining to "knock anyone who is doing it differently." - Improve: "slop" is borrowed as a community term and carries judgment the surrounding sentences work to avoid. #### Neurodiversity support - Present: short, one idea, conversational, with a transcript for people who prefer reading to listening. - Improve: none significant; the format is accommodating. #### Representation - Present: a single voice, which is the show's stated premise rather than an oversight. - Improve: the claim extends to "my hosts on all the podcasts we have," but none of those hosts speak for themselves here. #### Psychological safety - Present: the refusal to moralize is the piece's best quality — it states a practice without demanding others adopt it. - Improve: the absence of any accommodation framing means a creator who uses AI for accessibility reasons hears themselves lumped near "slop," which the host almost certainly does not intend. ### ✅ Logical Validity #### Strong reasoning - **The central distinction is well-drawn.** Separating AI that processes finished human work from AI that originates work is a real conceptual boundary, and it produces a rule that can actually be applied. - **Admitting AI use strengthens the claim.** A blanket denial would be less credible and probably less true; specifying where AI is used makes the commitment checkable in principle and more plausible in practice. - **The non-superiority disclaimer is load-bearing**, not decorative — it keeps a claim about one's own practice from becoming a claim about everyone else's. - **Scope is stated honestly:** he says he can "pretty much" say it for the other hosts, hedging rather than asserting flatly on their behalf. #### Fallacies - **Circular self-verification.** *"this was not scripted by AI. I actually wrote this script down and did it myself"* cannot distinguish itself from the thing it denies. An AI-written script would make the identical claim. Fix: nothing available today short of a provenance mechanism — which is itself the finding. - **Unsupported assertion throughout.** Zero external links, zero evidence, verified against the markup rather than assumed. Every factual claim rests on the speaker's word. - **Claim made on others' behalf.** "My hosts on all the podcasts we have" is asserted without any of those hosts present, though the "pretty much" hedge partly covers it. - **Undefined boundary term.** "Slop" does the work of marking the other side of the line and is never defined, so the line's exact location stays vague. - **Title/content mismatch.** "And the Open Web" promises an argument the episode never makes — and the page's own `<title>` tag carries a different headline again. #### Tone of voice - Warm, unpolished, deliberately conversational — filler words retained, one audible self-interruption ("I'm pinching myself now. Well, that's kind of weird"). The register is the argument: an over-produced version of this episode would undercut its own claim. It stays clear of moralizing through two explicit disclaimers, and the closing note about still liking to write is self-deprecating rather than triumphant. The one place the tone hardens is "slop," a borrowed epithet in an otherwise generous piece. ### 🛡️ DARVO Analysis DARVO describes a sequence by an accused party — deny the conduct, attack the accuser, reverse the victim and offender roles. **It does not apply here.** No one has accused this host of anything, no dispute is in play, and there is no accuser present. What is worth naming accurately instead: - **Deny.** There is a denial — that the content is AI-generated — but it is unprompted and pre-emptive rather than responsive. Volunteering a disclosure nobody demanded is close to the opposite of the defensive posture DARVO describes. - **Attack.** Explicitly declined, twice. The host states he is not boasting and not "knock[ing] anyone who is doing it differently." The only edge in the episode is the word "slop," which is aimed at a category of output rather than at any person. - **Reverse victim/offender.** Not present in any form. No injured party is claimed and no blame is relocated. - **Worth noting instead.** The structure actually present is a pre-emptive authenticity claim in a market where such claims are becoming unverifiable. The analytically interesting question is not whether the speaker is defensive — he is not — but that sincerity and performance of sincerity have become indistinguishable from the artifact alone. That is a property of the medium in 2026, not a fault of this speaker. ## 🎯 Conclusion A three-and-a-half-minute episode with one idea worth extracting. The host does not claim to avoid AI; he claims a boundary — AI may work on finished human material, and may not originate it. *"That is simply AI doing something useful with content a human already made. It's really not the same as AI making the content in the first place."* As disclosure language that is more precise and more honest than most, and it generalizes past audio to anything published. The limitation is structural rather than personal. Nothing here is verifiable. No links, no evidence, and a closing self-attestation that an AI-written script could make word for word. The episode does not pretend otherwise, but it also does not reach for any of the provenance mechanisms that exist, and a listener is left with a choice to trust rather than a means to check. Two smaller notes. The page publishes an unfilled `[Guest Name]` placeholder where the human's name should appear — a CMS oversight, mildly unfortunate on this particular page. And the title's promise of "the Open Web" goes entirely undelivered; the audio never touches protocols, syndication, or platform independence. **Next steps for readers.** Take the distinction and write it into your own disclosure, where it will do more work than a blanket denial. Then notice what the episode makes visible without arguing: that in 2026 an honest authenticity claim and a fabricated one are indistinguishable from the artifact, and that the fix is tooling nobody has adopted yet. ## 🔄 Refactor Appendix — Why Authentic Voices Matter in Podcasting and the Open Web ### 📚 Literature Atoms (H4) #### [H1] Post-processing with AI differs categorically from AI origination > [!quote] That is simply AI doing something useful with content a human already made. It’s really not the same as AI making the content in the first place, or what a lot of people call slop. Where: the episode's central passage, after the host specifies his own AI use for show notes, takeaways and derived posts. Why it matters: the most usable disclosure framing in the piece and the reason to keep it. "Did you use AI" is losing diagnostic value; "did a human originate this" retains it, and this draws the line cleanly enough to apply. <!-- atomicity: 1 idea ✓ --> #### [H2] Admitting partial AI use strengthens rather than weakens a provenance claim > [!quote] I do use AI. I’ll be straight about that. Where: immediately preceding the distinction. Why it matters: a blanket denial would be less credible and probably less accurate. Specifying where AI is used converts an unfalsifiable slogan into a scoped commitment that could in principle be checked. <!-- atomicity: 1 idea ✓ --> #### [H3] Human origination is the constant the host claims across format change > [!quote] There isn’t a script or an idea that started as a prompt. Where: the claim covering eight years of the site and the network's shows. Why it matters: states the commitment at the level of origination — idea and script — rather than output, which is the right place to draw it and the hardest place to verify. <!-- atomicity: 1 idea ✓ --> #### [H4] A self-attested authenticity claim cannot distinguish itself from its counterfeit > [!quote] this was not scripted by AI. I actually wrote this script down and did it myself. Where: the closing line. Why it matters: recorded not to doubt the speaker but because the sentence is structurally identical to what an AI-written script asserting human authorship would produce. It is the clearest small illustration available of why provenance needs mechanism rather than assertion. <!-- atomicity: 1 idea ✓ --> #### [H5] The claimed differentiator is presence, not production quality > [!quote] It’s not a format, not a workflow, just me talking. Where: the episode's closing positioning. Why it matters: names what the host thinks is actually scarce — a person accountable for the view being expressed — as distinct from anything about how the audio was made. That is the competitive claim independent creators are increasingly making, stated plainly. <!-- atomicity: 1 idea ✓ --> ### 🧠 Permanent Candidates (H4) #### [P1] Scoped disclosure is more credible than blanket denial Admitting where a tool is used and naming the boundary produces a commitment someone could test, where a categorical denial invites disbelief precisely because it is unfalsifiable and probably untrue. <!-- atomicity: 1 idea ✓ --> #### [P2] An authenticity claim without a verification mechanism conveys intent, not fact Where sincerity and its imitation produce identical artifacts, a stated commitment tells an audience what the speaker wants to be held to and nothing about what occurred. <!-- atomicity: 1 idea ✓ --> ### ✏️ Fleeting Queue - #xref The origination/post-processing split is directly adoptable as a disclosure line for Courtney's own publishing, including AI-assisted processing of sources she selected and read. - #articleseed What a practical provenance mechanism for independent audio would look like, given that C2PA-style signing exists and nobody uses it. - #xref The title promises "and the Open Web" and never delivers; the Standard.site guide read earlier today is the argument this episode was missing. - #xref No accommodation framing — creators using AI to lower a language or disability barrier sit uncomfortably near "slop" here, almost certainly unintentionally. - #aphorisms "It's not a format, not a workflow, just me talking." ### → Things 3 (max 2 actions) - [ ] Write the origination-versus-post-processing distinction into a reusable AI-disclosure line for published work — context #things - [ ] Read this against the Standard.site note for the open-web argument this episode skipped — context #things ``` Processed: Why Authentic Voices Matter in Podcasting and the Open Web - Literature atoms: 5 / 5 cap - Permanent candidates: 2 / 3 cap - Fleeting items: 5 - Things 3 actions: 2 / 2 cap - Konik tags used: #xref, #articleseed, #aphorisms - Triage callouts populated: Cross Reference, Article Ideas ``` <!-- deep-read:sections end --> 📄 **Full text:** [[Full Document Contents/Articles/Why Authentic Voices Matter in Podcasting and the Open Web.md|Why Authentic Voices Matter in Podcasting and the Open Web by Bob (host, Open Signal)]] ## Highlights - That is simply AI doing something useful with content a human already made. It’s really not the same as AI making the content in the first place, or what a lot of people call slop. ([View Highlight](https://read.readwise.io/read/01m28b0nm3xwt4ytm38f904v7w)) - Note: The reusable disclosure line. "Did you use AI" is losing diagnostic value; "did a human originate this" retains it. Transfers to writing, video and code. - Tags: aphorisms, ai-disclosure - this was not scripted by AI. I actually wrote this script down and did it myself. ([View Highlight](https://read.readwise.io/read/01m28b0rgvcyx7h65280qh8vvz)) - Note: Recorded not to doubt him but because the sentence is structurally identical to what an AI-written script claiming human authorship would produce. The clearest small case for why provenance needs mechanism, not assertion. - Tags: content-provenance, xref