Podcasting 2.0

Podcasting 2.0 October 2nd 2026 Episode 273 - "Advertising is The Horse and Buggies of Podcasting"

One topic only today - Big changes coming to podcast discovery!

05 - πŸ€– PODCAST ROBOT — ADAM'S MCP SERVER FOR THE INDEX

Discovery

New Tag adoption

podroll trailer peson location etc

AEO - Ask Engine Optimization

AI Assistants - Robots - Muse

podcastindexrobot.com — Podcast Robot, part of Podcast Index

What it says it is: "Podcast Robot is a plug-in for AI assistants like Claude, ChatGPT or your own agent. Once connected, your assistant can see every podcast in the world through Podcast Index… plus the live signals that say when something new drops or a show goes live. No account, no key, nothing to install. Your podcast app stays your podcast app: Podcast Robot hands you links that open right in it."

The asks it handles: "What's new in my shows?" · "Find the part where they talk about X" (transcript search, link to the moment) · "Who was that guest? Where else is she?" (podcast:person) · "Surprise me" / what other podcasters recommend (podroll) · "What's made in Austin?" (18,000 shows carry location) · "How do I support this show?" (funding / Value4Value).

Let it keep watch: "Tell me when [show] goes live" · "Watch for any episode where [person] is a guest" — it tells the assistant to check only what changed, so one check costs one or two calls. The page is honest: hosted assistants schedule hourly at best (ChatGPT, Claude Routines, Dots); a true every-15-minutes needs your own agent, such as Muse.

πŸ’‘ For podcasters: "grade my feed" — Podcast Robot "tells you which Podcasting 2.0 tags you are missing and what each one would let assistants do for your listeners." That is an answer to Sam's marketing thread in section 01: not a version number, a reason — every tag is now something an assistant can do.

Its links out use podcast-platform-links (GitHub)

csb's prediction this morning: "AC today on Podcasting 2.0 will say: Podcast player app developers, please abandon your apps because future of Podcasting are: Always-On AI Agents." The page says the opposite: your podcast app stays your podcast app.

Podcasting 2.0 September 25th 2026 Episode 272 - "Is this Y'all?"

00 - πŸ—„οΈ THE SCHEMA IS LIVE — AND THEN THE AGGREGATORS PICKED A FIGHT

Dave has already told us this one is his: "Will explain in the board meeting tomorrow."

Mon 22 Sep — the cutover lands: "#api All api endpoints are now live with the new database schema. Bug reports welcome. I'm tired. πŸ˜…"

@dave — all endpoints live on the new schema

Tue 23 Sep — and now the real work: "Now comes the post-deploy quality testing of the Index as a whole. Now that we have an audit trail for every feed we will need to walk every podcast in the Index to see where it stands and what metadata needs reconciling."

New tables in the dump: "@js There are new tables in the ndjson dump now. πŸ˜‰" · "@js You're the guinea pig. Let me know if anything looks off."

On breakage, to Mitch (Podverse): "Any additional parameters should be additive. The overall shape of the response should not change. There *shouldn't* be any breakage unless the client is doing a Rust style strict type coercion. But, I wouldn't expect that to be common since JSON is fragile."

Thu 24 Sep — the brawl, in four posts:

12:10 — "#api Adding 2 new indexes on the feeds table. Expect the api to choke for a few minutes until this completes."

14:38 — "#api All out brawl with the feed aggregators this morning. I'm about to win though."

17:08 — "Battle is over. The war is ongoing."

20:02 — "#api I fixed a huge problem with the feed aggregators. Based on this snapshot of our database metrics, can you guess at roughly what time that happened?" · 20:03 — "Network throughput up. CPU and disk rate down. It's a win-win."

πŸ” The question for Dave: what were the aggregators actually doing — hammering the feeds, or something subtler that the new audit trail made visible for the first time? And what did the two indexes fix?

19:13 — the bug he flagged for today: "#api There was a problem with search ranking in the search/byterm endpoint after switching the db schema. This should be fixed now. Will explain in the board meeting tomorrow."

Loose end: @ChadF, 24 Sep — "@dave looks like podcaster wallet is returning the old node address." Not visibly answered yet.

Also closed this week: docs-api #160, "BUG: Wrong season number in some episodes" — season 14 coming back as 4.

-------------------------------------------------------------------------------------------------------------------------------------

01 - 🎀 GUEST: SAM SETHI — TRUEFANS, AND THE THREE FEEDS HE ASKED TO TALK ABOUT

He set this agenda himself. 22 Sep, to Dave: "thank you my friend. When are we going to get you back on PWR and equally I would love to come back on P2.0 to talk about SuperFeeds, Publisher Feeds and Network Feeds."

@samsethi — the ask (22 Sep)

πŸ“¦ 1. SUPERFEEDS — what he has actually shipped

His own definition, 19 Jul: "Traditional RSS feeds were designed to distribute podcast episodes, with the audio attached using the enclosure tag. Then came Podcasting 2.0, introducing more than 30 new namespace tags… A SuperFeed lets you combine much more than audio or video. Alongside your podcast episodes you can include blog posts, publisher updates, events, premium paid content — all delivered through a single RSS feed."

The bit that matters for the Index: SuperFeeds started (Jul 2024) as "the ability for a podcaster to add any new Podcasting 2.0 tag to their existing RSS feed and then publish the new TrueFans RSS feed to the Podcast Index while staying with their existing host" — a free claim-and-enrich service that keeps the enclosure with the original host.

Where it is now: two experimental tags added Apr 2026 — podcast:pinned and podcast:store — plus podroll and publisher feeds, with events and merch built on the publisher-feed pattern. "Our goal is to allow creators to build a SuperFeed by picking what elements are added to their RSS feed."

And the anti-Patreon argument, 7 May: "Apple's proprietary API for HLS video doesn't use RSS. TrueFans is going the other way. We are building Super RSS feeds with Audio, Video and Freemium/Premium RSS in the same feed. No need for Patreon private feeds."

πŸ” Ask: if a SuperFeed is one feed containing podcast + blog + events + merch + premium, what does an app do with the items it does not understand? Is there a medium or a filter, or does every app have to learn every type?

πŸ“‘ 2. PUBLISHER FEEDS — and the pattern he keeps reusing

He builds everything on remote items off the publisher feed: "We will add NEW Events and Merchant tags based on the publisher feed" — `<podcast:publisher><podcast:remoteItem medium="publisher" feedUrl="https://feeds.truefans.fm/rss/publishers/…" /></podcast:publisher>`

Event feed proposal, 1 Aug: "We are proposing a new EVENT feed based on the Publisher Feed" — events in the admin dashboard appear on the podcast page; tickets sold externally or in-app via Apple Pay / Stripe. Live example: the Cheerful Earful Comedy Festival on one page.

Merch — the StoreFront proposal, 31 Aug: "TrueFans recently added a new experimental tag for RSS feeds called Podcast:StoreFront. The aim of this StoreFront tag is to mirror the Funding tag" — a URL to a third-party store (Amazon, Sticker Mule, Shopify) puts a shop icon on the page. Later: products hosted in TrueFans, paid in fiat via Stripe or in SATs.

πŸ” Ask Dave: is the publisher-feed-with-remote-items pattern the right substrate for events and merch, or is this three tags that each want their own medium?

πŸ•ΈοΈ 3. NETWORK FEEDS — the one nobody has seen yet

This is new. He has posted publicly about SuperFeeds since 2024 and Publisher Feeds since 2024 — but "Network Feeds" appears exactly once anywhere on podcastindex.social: in his 22 Sep ask. There is no proposal, no example feed, no thread. This is the scoop of the segment — let him define it live.

πŸ” Ask: what is a Network Feed, and how is it not just a publisher feed with more members? Who owns it — the network or the shows? What happens to a show that leaves?

-------------------------------------------------------------------------------------------------------------------------------------

02 - πŸ’³ SAM ON PAYMENTS — "THE FUTURE IS MICROPAYMENTS… BUT MAYBE NOT BITCOIN"

🚨 This is the live wire of the episode. V4V was built on Lightning and SATs. Our guest — who has shipped a SATs wallet — is now saying the rail may be wrong.

24 Sep, in his own words: "We also just enabled Paypal, Amazon Pay, Samsung Pay to go alongside Apple/Google Pay and Bank. I am now looking at Stablecoin as an option. I still [think] the future is micropayments direct from fan to creator but maybe not Bitcoin."

Stripe Pay By Bank — TrueFans added support same day: "Pay by Bank is a single-use payment method that allows customers to pay directly from their bank account instead of using a card… runs on banking infrastructure and takes advantage of open banking APIs." Caveat he flagged: "only Finland and UK which is odd but we will keep an eye and see if other countries get added."

25 Sep 00:29, the thesis: "there is still a real need on the web for micropayments. I think when mainstream banks give customers a current/chequing account and a digital account and make it easy for us to transfer money from one account to the other will be the start of a wider adoption."

27 Jul, the post-mortem on v1: "Maybe v1.0 was too complex with KYC, Nostr, Alby and Moonpay. I am watching Stripe, Stablecoins and x.402, a new open payment protocol that makes internet-native micropayments…" — and the Dizzee Rascal quote: "Some people think I'm bonkers / But I just think I'm free."

πŸ’‘ But his funding tag is the purest V4V in the ecosystem — give him credit for it: "The podcaster only gets paid if they continue to publish more episodes… Platforms like Patreon or Substack will continue to take the monthly amount even if the podcaster doesn't publish." The fan sets a monthly maximum; TrueFans draws from it per episode published. That is value-for-value enforced in code.

Also: if a feed has no funding tag, TrueFans turns its own on and takes 10% of funding transactions. Worth asking how that sits with podcasters who did not ask for it.

πŸ” The question for the room: if Pay By Bank and stablecoins do the job with less friction, is Lightning still the point of V4V — or was the point always fan-to-creator with no intermediary, whatever the rail?

-------------------------------------------------------------------------------------------------------------------------------------

03 - πŸ›οΈ SAM'S THESIS — "LESS LIKE PODCAST APPS AND MORE LIKE CREATOR PORTALS"

15 Sep, his four-part statement of direction: "The old model was simple: one creator → one podcast → one audio app. That worked when podcasting was primarily about listening. But creators are building something much bigger now… the next generation of creator platforms will look less like podcast apps and more like Creator Portals — a single place where creators can build, engage, communicate with and monetise their communities."

What is already live: host and stream audio or video, publish blogs, run live events, sell tickets, merch links, premium per-episode and per-blog pricing, a fan wallet.

New this month — newsletters: "Starting next week we're taking another step forward with the launch of the NEW TrueFans Newsletter… the plan is to make the same newsletter functionality available to every TrueFans creator" with subscriber / delivered / opened / clicked stats. "The goal is to give creators one place to build, own, manage and monetise their entire relationship with their true fans."

Roadmap he published: customised domains · new native iOS & Android apps · CarPlay, Apple Watch, Apple TV.

Features built on P2.0 tags — the good citizen argument (31 Aug): "I'm excited that we are starting to see apps building new features on top of Podcasting 2.0 tags… I'm now looking at every tag, trying to see if we can build a new feature on top."

• Frequency tag → show the NEXT publish date instead of the last one · • Live Item tag → PENDING / LIVE NOW / END status on the page, localised · • Person tag → aggregate every podcast where you are tagged into one dynamically-updating playlist on your profile · • Location → new location search built on @alberto's Radar.

Coming: "Our next big feature will be to use AI to grab 5 clips from the audio and/or video… and add those clips to the Podcasting 2.0 soundbite tag, as well as auto publishing those clips to the podcasters social media." πŸ’‘ Machine-generated soundbites written back into the namespace — that is a namespace question, not just a feature.

And the gamification layer: 30 activity verbs tracking fan behaviour, points to identify SuperFans, fans can OAuth their socials and auto-publish chosen verbs — including how long they played and percent completed, with a play-percentage slider.

-------------------------------------------------------------------------------------------------------------------------------------

Adam & Dave have their SCBA's on and go deep into index refactoring and ad loads!

ShowNotes

-------------------------------------------------------------------------------------------------------------------------------------

00 - πŸ”₯ STEP TWO OF THE SCARY STUFF — THE ENDPOINTS FLIP TO THE IDENTITY LAYER

Last week the table went in. This week the API started using it — one endpoint at a time, four days running.

Sat 12 Sep: "#api I'm running a script to sync up podcasts.id and newsfeeds.id so that when a new feed is born they are in sync. During the cutover to podcasts.id as the canonical form, there were about 11k feeds that drifted away from being in sync with their podcast because of a gap in the auto_increment values. This isn't strictly necessary, it just will make human review a little less cumbersome."

Mon night 14 Sep: "#api Making some final refinements to the new api endpoints before deploying. These will be the first ones using the new podcast identity layer. 😰"

Tue 15 Sep — the first flip: "#api The podcasts/byfeedurl endpoint is deployed with the first changes to the new podcast identity layer. We will do these one at a time, watching for problems. You will see nothing different in the response shape under normal circumstances. The response 'feed' object's `id` property now means the podcast ID. The `feedId` property now means the underlying feeds table row that podcast is treating as canonical."

@dave — byfeedurl flips (15 Sep)

"It's not ideal to have the podcast id be represented as feeds.id in the new responses, but anything else would have broken existing apps. And, the meaning is still faithful to what it was before. That was always meant to be the ID of the podcasts. We just didn't have a way to logically express that."

Same day: "#api The new podcasts/byfeedid endpoint is deployed. That's all for today. Will watch the logs until tomorrow." · "#api First run of the itunes ID 404 reconciler under the new podcast identity regime. 🀞"

Thu 17 Sep: "#api The podcasts/byguid and podcasts/batch/byguid endpoints are now flipped to the new db schema." · Fri 18 Sep, this morning: "#api The new podcasts/byitunesid endpoint is live."

@dave — byitunesid live (18 Sep)

πŸ“ The app developers noticed. Mitch (Podverse): "listening to last week's P2.0... the nextgen Podverse schema (and ogen) relies heavily on the Podcast Index IDs as basically the authority on feed uniqueness. It sounds like we may need/want to do a substantial rewrite to account for how Podcast Index API nextgen will handle ids?… after it is settled, I'd appreciate any documentation."

@mitch — the Podverse question (15 Sep)

Dave: "If I do my job right, you will see nothing different and will not need to do anything. I made sure the starting point for every canonical podcast index 'podcast id' is the current live feed id. If you see some sort of drift or problem it would be a bug I need to fix… the existing podcast ID's have always identified feeds, but now they will identify podcasts. The ID's themselves will not and have not changed. I'll definitely be producing docs to flesh this out further though."

πŸ˜„ Then the drawing: "It's actually kind of confusing to explain. It's easier if I draw it out on paper. It sort of needs to be visualized." → "#api These might be very hard to follow. There are two 'entrances' into the redirection chain and the Index follows whichever one you come in on." → "Paper and pen was faster than some AI agent."

@dave — the two entrances, drawn by hand (15 Sep)

The rule as Dave wrote it: "If you come in on the 'byid' path, you start at the podcast layer and follow the redirection to the correct feed and serve it. If you come in on the 'byurl' path, you start at the feeds layer and follow up the chain to the podcasts layer and then follow the redirection to the current canonical feed and serve that."

And the MCP ask, parked: @csb asked for an MCP server. Dave: "Have to get through the rollout of the new schema first. Adding MCP now would just be too much confusion. It's on the list for after all the new endpoints are delivered."

⭐ Board questions:

1. Four endpoints flipped in four days with nothing caught fire. Which endpoints are left, and which one are you most worried about — search? episodes? podping?

2. Walk the two entrances on air. If I hit byid with an old feed ID that has since been merged into another podcast — what comes back? The new canonical feed, silently? A redirect flag? Nothing?

3. Mitch's question is every app developer's question: when do the docs land, and is there a test endpoint where a developer can throw a redirected/merged feed at it and see the new behaviour before it hits their users?

4. The itunes-ID 404 reconciler — what does it do now that a podcast can outlive its feed? Does a dead iTunes ID move to the podcast row or stay with the feed row?

5. The 11k drifted IDs — human review "a little less cumbersome." How many merge decisions are sitting in the conflict queue, and who is the human?

-------------------------------------------------------------------------------------------------------------------------------------

08 - 🐧 OMARCHY, AND THE 'AI SUMMARIZER' EVERY PODCAST WEBSITE NOW HAS

Dave went Arch. Sat 12 Sep, to Aleck Gates ("Why is everyone using Omarchy? Did I miss something important? My Fedora desktop just keeps chugging along"): "There is no doubt some of it is the DHH spell some people are under. I find him irritating so my draw was just having an easy starting point (config-wise) into Arch… I also wanted Hyprland and after I went to Ubuntu 24 on that box I saw Hyprland maintainers recommend against using it on Debian distros. It came down to Arch or Nix after that." (You told the NA audience on 1901 to give the kids an old laptop and "install Omarchy.")

@dave — the Omarchy thread (12 Sep)

@csb's Omarchy facts of the week: "Lex has just reviewed Omarchy Linux creator" · "uses Slavic-made font: JetBrains" · "creator is EU citizen creating Omarchy in EU, you are welcome, Americans" · "can be installed from USB stick under 60 seconds." @jaybrunet shipped an llm-insert plugin for the micro editor — "non-destructive, precise editing of config files when you don't want to send your whole config to a model on OpenRouter."

πŸ“œ Dave Winer, 11 Sep 2001, reposted by JS on the anniversary: "Someday soon every home will have a weblog, and we'll have great aggregation tools that allow us to quickly assemble lists of loved ones who survived. A new button on cellphones that says 'I made it' and it flows the fact to all your concerned friends."

scripting.com — 11 Sep 2001

-------------------------------------------------------------------------------------------------------------------------------------

03 - πŸ—½ PODCAST WEEK NYC — WEBSTER'S 'FRONT OF THE STORE', THE IAB SAYS NO NEW FRAMEWORK, AND A 99-SECOND PNWR

The week: Sounds Profitable Business Summit (Mon–Tue, 400+ people, 200 buyers), the IAB Podcast Upfront (Wed), Podcast Movement (Thu–Fri). Podnews was there all week; PNWR was 99 seconds.

πŸ›’ Tom Webster's keynote — 'Are podcasts at the front of the store?' Previewing The Podcast Landscape 2026: "for the first time, the top three places where people have 'ever' consumed podcasts are not dedicated podcast apps. YouTube, Spotify and Netflix don't show podcasts front-and-centre… Webster compared Netflix's podcasts category as the equivalent of the CD range in a Walmart — just the most popular titles in a small selection which is hard to find; while a real podcast app like Apple Podcasts or Pocket Casts was the equivalent of Tower Records." Rob Walch, separately: "Netflix lists just 1 in 67,547 podcasts. (And doesn't list yours.)"

Podnews — Are podcasts at the front of the store? (18 Sep)

πŸ“ The IAB: a new framework is 'not needed'. Jason Adamski at the Upfront: the IAB "appears to reject a medium-agnostic measurement standard like the recently-released AMP Accords, saying 'the solution is not to build a new framework from scratch', and instead suggests… two separate standards from the IAB Tech Lab — the Podcast Measurement Technical Guidelines v2.2 and the Digital Video Served Impression Guidelines." Cridland: "The guide — and even the full press release for it — requires a registration with your email address to read. That's not how you set standards." He hosts the PDF himself.

Podnews — IAB: a new framework is 'not needed' (17 Sep)

Dave asked the question on the instance (16 Sep): "@samsethi Do you still have to pay to see the actual spec?" James: "You've never had to pay to see the spec. (This isn't v2.2 — which has been out for a few years — just a guideline that as far as I can see rejects the AMP Accords entirely)." Dave: "I could have sworn there was something you had to pay for. The IP block lists or something maybe?" James: "Certification and I think the IP blocks as well…"

πŸ“‹ The Podcast Study 2026 (Summit): "45% of podcast listeners would be 'at least somewhat likely' to pay for a subscription tier to remove ads; and 64% would sign up for email from their favourite podcasters." Edison: topic, hosts and guests decide what people pick; trailer and cover art least.

🎬 Video, still not a platform: Podscribe at the Upfront — "not one of the top 500 shows are available on all video platforms: YouTube, Spotify, Rumble and HLS via open RSS." Amazon Music adds ART19, Triton Omny and Flightcast video (via the alternateEnclosure tag, HLS), six countries. Podscribe–Rumble attribution is "the first video platform to support all the signals required… under the AMP attribution specification."

πŸ“» Machine Gods: Kevin Roose and Casey Newton leave the NYT's Hard Fork for NPR — a video show from October, on member stations from next year; the first deal under NPR's new content chief. ("We guess they like tote bags.")

Podnews — A new home for Roose and Newton (16 Sep)

πŸ’° Money moves: Substack adds a podcast tab — creators "collectively earn over $200 million per year on Substack"; Spotify Partner Program expands to 35 more markets, "total monthly payouts to enrolled shows have increased by more than a third"; Megaphone quietly raised SPAN's bar to three published episodes, undocumented, no comment; Giggly Squad's SiriusXM renewal "multiple offers around $150 million"; a Disctopia patent that checks whether you are listening by asking you questions mid-episode, "part of any payment to creators."

Podnews — Substack podcast tab, $200M a year (11 Sep)

πŸ›  Tooling: Podpage Agent-Built Websites — "work with an AI agent, like Claude or Codex, to design your websites; and then use Podpage to make it properly hosted"; PodAnalyst v3 lets you "connect Iris to your own Claude or ChatGPT account"; Headliner auto-clips for YouTube; Threads now unfurls Apple/Spotify/iHeart podcast links; the Podcast Standards Project met Thursday 12:30 ET near the venue, open to non-members.

🏷 Our people at PM: Mike Dell (Blubrry): "I'm headed there too. Speaking about podcasting 2.0 stuff." Matt Medeiros was going "bearing Podcasting 2.0 gifts — I'm advocating for the funding tag… Forget views — get the funding tag!" then: "Sadly I had to cancel my trip to podcast movement. But I'm thinking of doing a virtual event around fundingtag.link." Sam Sethi: "we featured you on Creators from TrueFans and talked about the three steps you need to undertake to make the Funding Tag work."

fundingtag.link

Sam Sethi's thread (15 Sep), the Creator Portal: "The old model was simple: one creator → one podcast → one audio app… the next generation of creator platforms will look less like podcast apps and more like Creator Portals" — TrueFans newsletter launching next week, subscriber/open/click data to creators; roadmap: custom domains, native iOS/Android, CarPlay/Watch/TV, stablecoin support.

@samsethi — the Creator Portal thread

⭐ Board questions: (1) Webster's Walmart-vs-Tower-Records: is the Index's answer to Netflix a better front of the store — newpodcastapps.com — or is it that we never wanted to be in Walmart? (2) The IAB keeps its guide behind an email wall and its certification behind a fee; the AMP Accords are open. Does the Index endorse a measurement spec, or stay out? (3) Substack's $200M and Spotify's payouts are both closed-loop money. Where is the V4V number for the week — stats.podcastindex.org/v4v.

-------------------------------------------------------------------------------------------------------------------------------------

Podcasting 2.0 September 11th 2026 Episode 270 - "YOLO Guy

Dave waves the whitflag on agentic coding and podcasting beneifts! 

01 - πŸ˜‚ THE PODCAST RIG ATE ITSELF — AND IS NOW RUNNING OMARCHY

Sunday 6 Sep: "My Umbrel docker containers did not survive the OS upgrade from Ubuntu 20 to 24 on my podcast rig. This is why I never upgrade anything. I still have some Ubuntu 12 and 18 servers running." — a sentence that should be on a T-shirt at Podcast Movement.

Tuesday, the diagnosis, and he takes it on the chin: "It was a v1 to v2 docker compose transition issue. It was very much my fault for not RTFM before doing do-release-upgrade." Plus: "doing 2 upgrades back to back to get from LTS 20 to 24."

To @ericpp, same day, the two-line eulogy: "It's incredibly borked." → "I'm closing the channel you have open to me because I'm going to have to wipe this box." — note that the lightning channel was collateral damage. That matters for §02.

Thursday 10 Sep, the rebuild: "Omarchy on the podcast rig now. Let's see how this goes." and then, shortly after, "Podping gossip watcher up and running on it."

The genuine question underneath: the podcast rig runs podping's gossip watcher. What is actually load-bearing on that box, and should anything that the wider ecosystem depends on be sitting on a machine you're willing to wipe on a Tuesday?

-------------------------------------------------------------------------------------------------------------------------------------

09 - πŸ€– THE ANTHROPIC ADVENT CALENDAR

@js to you directly, 11 Sep: "@adam it's like an advent calendar until the IPO — behind today's door is..." and then the Anthropic threat-intelligence report.

Anthropic — Detecting and countering misuse (11 Sep PDF)

The line JS pulled out: "These actors carried out a sustained effort to develop guided weapons, using Claude to design guidance software. They did test-fire a guided rocket. This field test appears to have failed: within hours, the ac[tors]..."

And the prediction, two days earlier: "can you _imagine_ what stunts anthropic has planned for around the IPO timeline in a few months? gonna be great."

The board-meeting angle: every one of these reports is simultaneously a safety disclosure and a capability advertisement. Publishing "our model was used to design a guided rocket" is a warning and a brochure at the same time — and the IPO is in a few months.

-------------------------------------------------------------------------------------------------------------------------------------

00 - πŸ”₯ STEP ONE OF THE SCARY STUFF — THE NEW 'PODCASTS' IDENTITY TABLE

The lead, and it's Dave's. Wednesday 3 Sep: "#api The new `podcasts` table is now being written to by default on all new incoming feeds and the conflict queue is being populated for merge decisions. This is step 1 of the scary stuff. Let's hope nothing catches fire."

@dave — the podcasts table goes live (3 Sep)

Thursday 4 Sep, step one lands: "#api All feed ingestion paths now use the new 'podcasts' identity table. Spent the morning cleaning up a lot of old error log debris I've been neglecting for years so I can have clean logs for the next step."

πŸ˜„ And the same morning he broke it and immediately confessed: "#api I would like to pretend (by silence) that I didn't break feed ingestion this morning for about 20 minutes by forgetting to update the env files on the aggregator servers. But, I'm sure @js will produce s[creenshots]..." — Dave knows exactly who is watching.

Tuesday 9 Sep, the number of the week: "#api Refactored the weekly database dump script since it was getting slower and slower and was very old. Re-running the new one now to test. I'm anticipating 4 hour run instead of 28 hours." — twenty-eight hours to four. That is the dump the whole developer community pulls from.

Podcast Index — public datasets

⭐ The board questions, and they are the show:

1. What IS the identity table — in plain language, for an app developer? If a feed changes URL, changes host, or gets duplicated, what does the new table let the Index say that it couldn't say before?

2. The conflict queue is "being populated for merge decisions" — who makes those decisions, and is it automatic or does a human rule on it? Because merging two podcasts that are actually one podcast is exactly the kind of call that goes wrong loudly.

3. You said step one. What is step two, and is that the part where something does catch fire?

4. Twenty-eight hours to four on the dump — what changed, and does that mean the public dataset can now ship more often than weekly?

-------------------------------------------------------------------------------------------------------------------------------------

Shownotes

-------------------------------------------------------------------------------------------------------------------------------------

00 - DATABASE SURGERY WEEK — AND THE THIRD EMERGENCY OF THE MONTH

The lead, and it's Dave's: Saturday 22 Aug — "#api I'm in the middle of a big refactor with the Podcast Index database schema. If I do my job right, nobody will notice it. But, the side effect will be an overall huge increase in the consistency of the data (duplicates, url change lag, etc)."

@dave — the big schema refactor (22 Aug)

Wednesday night: "Podcast Index database surgery will begin in the morning." Thursday: "Step 1. Expect a brief api outage. It should last just a few minutes." — then "Done" — then "#api Adding a very large table index..." — and finally, Thursday evening: "Finished. Butt is intact." (6 favourites. The board's favourite status update of the month.)

What Adam should ask: what actually changed under the hood, what does "consistency" buy an app developer in practice, and is there anything a host or an app should do differently now that duplicates and url-change lag are being cleaned up?

πŸ”₯ The other half of the week — the third emergency. Tuesday 25 Aug: "#api Linode emergency maintenance on our db host. API is down." Then: "No update as of yet. Just hard down." Then the line that is really the story: "This is the third 'emergency maintenance' event that has affected us this month. Getting kind of old." And, dryly: "I guess a reboot every few years isn't too much to ask."

@dave — Linode emergency maintenance, API is down (25 Aug)

Board question: three unplanned outages in one month on the host that carries the Index. Is it time to talk publicly about where the database lives, and what the fallback is? The Index is infrastructure now — other people's products go down when it does.

Podping shipped twice. v0.13.0 across gossip writer, watcher and monitor — "a dynamic memory cap to address the Iroh bug" (that is last week's memory leak, now capped). Then v0.14.0 the same day — "cpu architecture in the mesh announcement messages to make determining what hardware a node is running on easier." Monitor updated too: ppmonitor.podcastindex.org.

@dave — podping v0.13.0, dynamic memory cap for the Iroh bug

Podping monitor

Podcast Index API docs

-------------------------------------------------------------------------------------------------------------------------------------

04 - πŸ“Œ PODCASTING'S GOLDEN OPPORTUNITY: HUMAN SPEECH

Your PREP note, verbatim: "For PC2.0 - Podcasting's Golden opportunity! Human speech!" — filed against a piece on how to spot AI writing.

Can you learn how to detect AI writing? Here are some clues

The argument: text is now cheap to fake, images are cheap to fake, and as of this year so is a voice. What is not cheap to fake is a specific human, speaking unscripted, for three hours, with a transcript and a signature chain attached. Podcasting is the only medium that ships that by default — and the namespace already carries the provenance parts.

πŸ”₯ Evidence from this week, and it is stacked:

1. Apple Music is about to start labelling. @js: "more than a third of the music Apple receives monthly is '100% AI' + usage — presumably what other humanoids would call 'listening' — of AI music on Apple Music is below 0.5%. = yes they are going to have to do something about it, starting with labelling." Read that ratio again: a third of the supply, half a percent of the demand.

Billboard — Apple Music to label AI-generated music

2. Nobody can hear the difference in podcasting either. @jaybrunet, 22 Aug: "I listened to a voice clone podcast last night ('Stark Reflections') and it was surprisingly more concise and easier to follow than his previous episodes that tend to meander off in too many directions. If he hadn't said it was ElevenLabs, I wouldn't have cared or noticed the difference."

3. The scale of the flood, measured. Marco Arment said Overcast transcribes about 30,000 hours of audio per day. @js checked it against his own index: "taking a look at my 'all new episodes published' internal db for last month, the world published at least 1.4m hours of new mp3 episodes for the month, or about 46k hours a day... so if your 30k hours are unduplicated then you have excellent coverage — the long tail is very long and, these days, full of generated audio."

@js — 1.4m hours of new mp3 in July, full of generated audio

4. The shame is gone. Stanley Druckenmiller, on his WSJ op-ed: "I'm not embarrassed by it... I write everything using AI now for the same reason I use a calculator when I do math problems. I don't know why this is relevant. My name is on the piece. It's my message."

Notus — Druckenmiller's WSJ op-ed

5. And somebody is already building the counter-stack. Japanese publishers' Originator Profile: "OP uses cryptographic digital signature technology to create tamper-proof credentials for websites. Each ID card confirms who published the content, whether professional organizations vouch for that publisher, and information about the publisher's editorial and ethical standards." — that is a person tag with a signature on it.

Nieman Lab — Japanese publishers and Originator Profile

⭐ The namespace question this asks: we have person (who), location (where), transcript (what was said, word-level, timestamped). That is three quarters of a provenance stack that no AI feed can forge cheaply. Do we need a fourth — a signed claim of human origin — or is the honest move the opposite: an AI-disclosure tag, because the liars won't use either one and the honest shows will?

The V4V angle: if a third of the supply is synthetic and gets half a percent of the listening, then attention is already voting for humans. Value-for-value is the only monetisation model that pays for that directly — you cannot boost a bot you don't believe in.

-------------------------------------------------------------------------------------------------------------------------------------

02 - THE LOCATION TAG, WEEK TWO — IT IS INSIDE PRODUCTS NOW

Last week we watched a dataset become a globe in 48 hours. This week the apps shipped it.

Podhome, this morning: "New feature: Add a location to your show and to your episodes, and modern podcast apps can put your podcast on the map. You can also see your show on @alberto's fantastic radar.rss.io." — note show AND episode level, which is exactly the gap James flagged on the globe.

@podhome — location on shows and episodes (28 Aug)

Podhome location docs

😊 The demand was already there — @rgblack316, 26 Aug: "@podhome I was just about to send you a feature request to add the Location tag and logged in this morning to find it already there. 😁"

TrueFans shipped search off it — @samsethi, this morning: "New location search for discovery based on @alberto radar now added to TrueFans. truefans.fm/search/"

TrueFans location search

😏 And @silas had it first and said nothing — 22 Aug: "Listening to the episode and hearing them talk about radar. I have had a map thing showing podcast locations for actually like two months. Both on web and on mobile. Not yet on tv. Would be cool if I were happy with it and not still a perfectionist. I'm in the car and I actually pulled over to write this."

...and he already solved the flaw everyone else is about to hit: "I ran into that shitty browser location thing and I solved it by, on the web version, just not giving you a search near me button. Only on the phone app because that is guaranteed to have gps and an actual location — not just wherever the heck a desktop computer seemingly randomly guesses where it might be located."

@silas — two months of a podcast location map

The granularity problem, now with the definitive example — James Cridland on PNWR this week: "You can put a building in there, a business in there, you can put everything from the state of Queensland to Brisbane City Post Office. If you do a podcast with a pastor — as you know, somebody who used to be on MTV with big hair does — then don't just put the city where you are, put..." (← that is you and Pastor Jimmy, on air, in Australia.)

⭐ Sam Sethi's line, and it is the best summary of the project anyone gave this week: "What we're beginning to see is all the hard work from the people in the podcast 2.0 community is beginning to bear fruit now, because I see them as building blocks. The person tag, the location tag, the transcript. All of these were fundamental building blocks that now you can start to build services around, which is what we've been waiting for. And of course, the lag was waiting for the data to be filled in once the tags were available."

Board question: the globe is channel-level. Podhome now writes episode-level. Does Radar (and the dataset behind it) pick up item-level locations next — and does the Index want a recommended granularity in the spec, so we stop getting podcasts located in an area 2.6x the size of Texas?

radar.rss.io — the globe

Podcast Index — public datasets

-------------------------------------------------------------------------------------------------------------------------------------

03 - THE PERSON TAG GETS AN ENDPOINT

Dave, 22 Aug: "#api The new episodes by person endpoint: /api/1.0/episodes/byperson?q=James+Cridland&max=5. It reports directly what the [item] level [person] tags dictate."

@dave — the episodes/byperson endpoint (22 Aug)

James wired it into Pod News, and said so on this week's PNWR: "a very simple, possibly too simple, but a very simple person tag search, which is now integrated into the Pod News website. The first time I integrated it I didn't do it very well, but it's much better now, because I've actually followed the spec."

πŸ˜„ And he name-dropped you on purpose: "if you're listening to Adam Curry — let's just say Adam Curry, just so that his AI bot gets him to listen to this bit — you can click the search next to Adam Curry and it will query the podcast index person endpoint and find all the other shows Adam Curry has been marked down as a person on." (Confirmed: the bot got you to this bit.)

The drawback he found: "there is more than one person in the world called Dave Jones" — no disambiguation. A person tag is a string, not an identity.

Dave's own proposed fix, 25 Aug: "I think adding a toggle to the person search to switch it from extracted text to person tag only would be good. Then, on the person-tag-only option, there would be a way to filter for certain attributes like the href, image, etc." — the href IS the disambiguator. Wikidata/website URL as identity.

@dave — toggle for person-tag-only plus attribute filters

Sam Sethi, on PNWR, gently claiming priority: "We've included this as a feature for two years, it's called My Podcast Work... If the item level supersedes the channel, and in podcast:person there is a guest tag, then if we pick up that you are a guest we'll label that as a guest rather than the host."

🎯 Board bait, straight from James on PNWR: "I would have expected Dave Jones to have done an MCP directly with the Podcast Index by now. And who knows, maybe even to have stuck that behind some form of a paywall. Given that Dave is working so much with AI tools in his day job these days, I would imagine he'll be pretty good at making it all work. So I'm kind of surprised why they haven't done that yet." — Dave, answer that one on air.

Related, same week: @csb on WebMCP — "an emerging browser standard that lets a website explicitly declare its features as structured, callable tools for AI agents. Instead of forcing an AI agent to guess page layouts, read complex DOM trees, or parse scripts." — the question underneath both: does an open index expose itself to agents, or does it get scraped anyway?

-------------------------------------------------------------------------------------------------------------------------------------

01 - πŸ”₯ V4V FAILED ON OUR OWN SHOW — THE ALBY / KEYSEND MESS

The post of the week, and it is a gut-punch. @csb, Friday morning: "@dave @adam bros, I tried, oh I tried, to send you $15.10 via bitcoin lightning but 4 splits failing including the main one, please fix it as it's embarrassing: if podcasting 2.0 value for value system fails for the main podcast of the podcasting 2.0 movement then it will fail for non-celebrity podcasters too, looky:"

@csb — 4 splits failing including the main one (28 Aug)

Dave, same morning, and he does not dodge it: "Ugh. I opened a new public channel yesterday with LNServer and our AlbyHub says it's offline. Idk why. I guess I'll have to open another one. [sigh]" Then: "I did test yesterday and it was working. Looks like it went offline overnight." Then: "New channel is opening now and the other one just came back online." And, at 11:39: "Ok, both public channels are online now."

@dave — our AlbyHub says it's offline (28 Aug)

@dave — both public channels are online now

πŸ”₯ But this is not a one-morning outage — it is the third independent report of the same class of failure this week, and the other two are older:

1. @algrid, 25 Aug — the specific technical finding: "I recently did some V4V testing and noticed that [email protected] receives bolt11 payments successfully, but keysend payments to it fail. You probably should check your channels configuration there or remove the getalby.com/.well-known/keysend/podcastindex data to disable keysend completely if Alby allows that."

@algrid — bolt11 works, keysend fails (25 Aug)

2. @ChadF, 22 Aug — found it independently from the app side: "I also don't think the PI Alby address in the feed accepts keysend and is causing failures every time I send a boost in boostmebitch.com. Which is strange since every other Alby address is fine. I do the well known lookup and do keysend if it's available for Alby addresses."

@ChadF — PI Alby address and keysend failures (22 Aug)

3. @podcastguru, 21 Aug — the monitoring gap, stated plainly: "Looks like our Albyhub was down for a month and I had no idea. 😫 Had to update the hub, re-open channels, I assume payments still worked. Just kind of annoyed how the hub can go offline and we don't get any type of notification."

@podcastguru — our Albyhub was down for a month (21 Aug)

⭐ The board question, and it is the show: we run a podping monitor that tells the world when notifications stop. We run an API status nobody has to guess at. We do not monitor our own money endpoint. Should the Index publish a V4V health check — a small bot that sends itself one sat an hour to every address in the value block and screams when one fails? Every app developer in the community would use it.

The channel side of the story: @ericpp opened a channel to the Index this week — Dave: "03011dfdeeb31113db15620fbc0bc102f23e7d09f952b60615642ae2107e7db594 is the public key." Then the LNServer dance: "It says I can only have either all public or all private channels so I guess I'll have to close our existing one first." → "the new public LNServer channel is open now with 2,000,000 sat capacity." → and this morning, after it went dark: "Well, at least we know why it's cheap now. Lol"

@dave — well, at least we know why it's cheap now (28 Aug)

πŸ’‘ And the standing idea this all feeds — @ChadF on your single-QR ask: "Adam mentioned having a single QR for boosts but that requires something like The Split Box that StevenB made. Easy if you run a node as a backend, but then you're taking a user's payments, splitting it and sending the split payments out, which is in the money transmitter area so I'm not touching it." His alternative: "when we update the [value] tag we add an entry for a single lightning address then apps like mine can offer a QR option and just make an invoice for that one address. That removes splits but it's no different than having a PayPal donation to a single account."

@ChadF — the single-lightning-address idea in the value tag

Namespace question that falls straight out of it: does the value block want an optional fallback address — one that an app can use for a QR, and that a sender can retry against when the split fails? Today, when a split fails, the money simply does not arrive and nobody is told.

It's always DNS or CORS - and how your right brain is gating bad robots

Show Notes

-------------------------------------------------------------------------------------------------------------------------------------

00 - THE LOCATION TAG — SPEC TO DATASET TO GLOBE IN 48 HOURS

The lead, and it's yours: Wednesday afternoon Dave posted — "There is a new locations data set on the public datasets page... The db tables are still back-filling since I updated the db and aggregators to support the new <podcast:location> tag format. After a couple of days it should be mostly complete." He cc'd @james, @alberto, @js and then @DavidMarzalC. That cc list is the whole story — within 48 hours all four had done something with it.

@dave — the locations dataset post (20 Aug)

The shipping trail: web-ui PR #608, "add locations export to public data sets," branch public_datasets → merged to master the same afternoon. This is the first public dataset built on the new location tag format rather than the old one.

Then somebody built the map. By Friday morning Alberto had radar.rss.io/#globe up — a spinning globe of podcasts made in or about places on earth, drawn from the free dataset, most entries carrying full OpenStreetMap references. James put it in front of the OpenStreetMap community directly: "One for excellent people using @openstreetmap."

James's review, and the two asks: "This is so cool. Really shows the power of the location tag. The globe — would be nice to add filters to that (made in vs about). Or maybe just colour the dots differently? The next trick would be some OSM lookups on the IDs to work out what the places are."

πŸ”₯ The flaw he found, and it's a good one: "One drawback is that some people have chosen 'Queensland' as a location, which is 2.6x larger than Texas." — the granularity problem in one line. A location tag is only as useful as the smallest box someone is willing to put themselves in.

Dave, on seeing it — and he is in Birmingham, Alabama: "Wow, this is incredible. I already see a podcast from Birmingham that I didn't know was based here." ❀️ ❀️ Dave, Friday: "It's beautiful. I can see this data powering so many cool app and platform features."

@dave — Wow, this is incredible / Birmingham (20 Aug)

@dave — It's beautiful (21 Aug)

The bug found by using it: @alberto and @james noticed the rel property missing from the export. Dave: "That column is in there but I had to add it. I bet I left it out when I copied over the episode level response build code." Then: "I found the issue. I had made the change but not deployed the new export script to production. The 'rel' property should be in there now." Also confirmed: channel-level only in that dataset, not episode-level.

@dave — that column is in there but I had to add it

@dave — channel level only / rel now deployed

πŸ˜‚ Community request of the week — @ChadF: "we need someone to add the FEMA regions to this so I can add FEMA region 3 to my podcast location." (Board: FEMA Region 6 is already broadcasting from the Heart of the Texas Hill Country.)

@ChadF — add the FEMA regions

Board questions: does made in vs about want to be two tags or one tag with a rel? Do we want a recommended granularity (city, not state)? And is an OSM ID lookup something the Index should do once, centrally, so every app doesn't do it badly?

Podcast Index — public datasets

radar.rss.io — the globe

web-ui PR #608 — add locations export to public data sets

James on Mastodon — one for the OpenStreetMap people

-------------------------------------------------------------------------------------------------------------------------------------

01 - DAVE'S WORLD — THE MEMORY LEAK, THE REFACTOR, AND A WEBKIT WIN

Podping / Iroh is eating RAM since DHT came out. @suorcd and @alecksgates reported it; @ericpp confirmed on a Raspberry Pi 4: "It's strange because it works fine for about a day and then ramps up the memory usage hard." Dave: "Previously it was v0.97. Maybe there was a memory use regression when I removed DHT/DTT. There were memory problems before that. I'll investigate."

@dave — memory use regression when I removed DHT/DTT

@ericpp — same on a Raspberry Pi 4

@dave — might have been an Iroh reload

The target, stated: "1 gig should be way more than I want it to need. I'd love it to run on less than 500 MB of ram. I'm in the middle of a PI pretty big refactor. Let me finish that up and I'll do this next on the list. Feel free to explore things yourself though if you are so inclined."

@dave — 1 gig should be way more than I want it to need

And the suspect: "I'm positive it's QUIC. I just don't know why." — with Dave asking what debugging would actually trace it. Board: this is last week's "DHT is out and podping is go" meeting its first bill.

@dave — I'm positive it's QUIC

πŸ† The WebKit fix landed. Dave, Friday: "It's merged!" — a WebKit bug where a CORS preflight hangs forever if the server answers OPTIONS with valid Access-Control-Allow-* headers but a body that never terminates. WebKit was gating the preflight on the body finishing; Chrome and Firefox conclude it from the headers. So this was a Safari-only hang: fetch() never settles, bytes stream and get discarded until the page gives up. Fixed by concluding the preflight from status + headers as soon as they arrive. (Reported against bug 320657, reviewed by Youenn Fablet.)

@dave — It's merged!

Why it matters here: every browser-based podcast app and every web player that talks to an API cross-origin was one badly-behaved OPTIONS response away from hanging in Safari, with no error to show for it. Worth asking Dave what he was building when he hit it.

Dave on the supply-chain attacks: "These supply chain attacks are going to bite vibe coders hard. I'm sure agents are installing packages like crazy all the time behind the scenes without any visibility from the user."

@dave — supply chain attacks and vibe coders

And the one-liner: Dave, over a screenshot — "Does this seem like a system you would let do your taxes?"

@dave — would you let this system do your taxes?

WebKit commit — CORS preflight hang, merged

WebKit bug 320657

-------------------------------------------------------------------------------------------------------------------------------------

04 - SPURLOCK'S DESK

πŸ˜‚ The callback to last week's show. @js on YouTube's own skip button: "youtube nearly always gives me a 'was this helpful?' dialog when I hit the skip ahead button. I don't usually give product feedback, but I make it a point to answer this one every single time: 5 out of 5 stars β˜…β˜…β˜…β˜…β˜…." — We spent all of 267 on Spotify's Skip Ahead. Google has been asking permission for theirs.

@js — 5 out of 5 stars, every single time

Google ordered to stop making side-loading hard. Judge Donato ruled Google added unnecessary "anticompetitive friction" and ordered it rectified — easier downloads of alternative Android app stores. Directly relevant to anyone shipping a podcast app outside Play.

@js — Donato on anticompetitive friction

Google is buying Spirit Airlines' data. "The data on the block includes Spirit Airlines' employee emails, Microsoft Teams messages, spreadsheets, and calendars, as well as marketing, productivity, and operations data. Google plans to use the data for product development and training its AI models." A bankrupt company's internal comms are now a training-set asset.

@js — Spirit Airlines data for AI training

The AI-host business, described without embarrassment (NYT): at Inception AI, "Everyone paused to admire Lila Walker, a new knitting- and craft-focused persona... who had a chunky wool sweater, long, wavy tresses of red hair... 'Man, I'm crushing on her,' [the] chief production officer said." (See also §05 — that company has just rebranded.)

@js — Lila Walker and the crushing-on-her quote

MrBeast as a product of the metric — Kevin Munger: "his videos do not have any point except to maximize viewership... He is a creator without creative vision, and is in fact better understood as a creation of audience feedback filtered through the YouTube Apparatus." Pairs with §03.

@js — a creation of audience feedback

The AI-authorship dodge (Nature): "What I reject is the leap from 'AI wrote the sentences' to 'the expertise and thought leadership behind them must also be artificial'." @js's response is the better take: "the 'I-am-a-fraud'-inducing moment for him was not in fact doing it, but getting caught by the nice editor. Why so defensive? Why aren't you proud of it? Don't bury it in a footnote, make Claude first author."

@js — the Nature piece

@js — make Claude first author

Dave Winer, on agents in production: "So I have no idea how useful these things could be if it destroys stuff in deployed projects with real users. FeedLand was down for over an hour, and stressed me out no end. Really fucked up."

@js — FeedLand was down for over an hour

Apple, per Spurlock: beta 5 and 6 still "slamming features into beta... not in fixes-removals-only mode yet." And on the new design language: "I feel like Apple has really jumped the shark — this 'new design' is not winning any medals." Also spotted: AppManagedFeatures, possibly the groundwork for restricting leased devices after missed payments.

@js — more churn than I expected (beta 5)

@js — Apple has jumped the shark

@js — AppManagedFeatures

Cursed news item of the week: "LinkedIn Has Accidentally Become a Dating Site — Despite Its No-Romance Rules."

@js — LinkedIn has accidentally become a dating site

Ars Technica — Google ordered to make alternative Android app stores easier

Reuters — court delays hearing on Google's purchase of Spirit Airlines data

NYT — AI podcasts, fashion, pop avatars

Kevin Munger — We Need to Talk

Nature — AI wrote the sentences

Scripting News — Winer on an agent taking FeedLand down

-------------------------------------------------------------------------------------------------------------------------------------

-------------------------------------------------------------------------------------------------------------------------------------

00 - THE SPOTIFY "SKIP AHEAD" BUTTON — THE BIG ONE

What it is: Spotify is testing a button that appears on the playback screen the moment it thinks you'd want to skip — an intro, a sponsorship message, or an entire ad break. It is bigger than everything else on screen. One tap drops you at the end of the break, back in the show. Premium subscribers only, selected markets including US and UK, and only when the app is in the foreground.

Who broke it: James Cridland at Pod News, 4 August 2026 — an exclusive. Spotify never announced it. Semafor and most of the trade press picked it up afterwards; Cridland noted on air this week that some of them ran it without credit.

The number that makes it different from a skip button: Cridland's arithmetic — skipping a typical ad break by hand is nine presses of the 15-second button, plus a correction tap when you overshoot. Skip Ahead is one. "Some people have said, oh, but you've been able to skip for years and this is no different. But a button that only appears when it's stuff that Spotify thinks you should be skipping is a start."

It skips other people's money: the button fires on third-party ads — Acast, Odyssey, the New York Times — and on ads inside Spotify's own shows. It even skipped Search Engine's read for its own paid tier, PJ Vogt pitching Incognito Mode. One half of Spotify sells advertisers reach; the other half hands listeners a button to leave.

Spotify's answer, and the trapdoor in it: Spotify told Cridland it is "not affecting ad delivery." True only if delivery means to the device. Cridland: "If ad delivery is to the listener's ear, which is actually what people are paying for..." — the download still counts, the ad still gets billed, the human never hears it. Spotify's other defence: the button merely shows where other listeners already skip. Cridland: "I'm not entirely sold on that, but if that's what Spotify say, I mean they wouldn't lie."

πŸ”₯ THE LINE FOR THE BOARDROOM — read the terms of service: "What's to stop them from going, you know what, people are using that skip ahead button, why don't we just turn that on automatically? I went to have a look at the terms and conditions of Spotify. There is nothing to stop them from doing that. Nothing at all. And let's be clear, that would destroy the ad-supported ecosystem that podcasting is."

Why the scale makes it existential: Spotify is 72% of all podcast downloads in the Philippines, 47% in Mexico, about a third in Australia, 75%+ in some markets. ~300 million paying subscribers plus 500 million free. Cridland on the smaller ad-strippers: "Podtastic will never be that big... they're not going to materially harm the podcast industry. This is." And it sets the precedent — if Spotify does it, every other app can say the same.

The YouTube comparison, and why it fails: YouTube also lets you skip some sponsor reads — but YouTube pays creators to be there. Spotify pays nothing for an RSS show to be on Spotify. Same behaviour, opposite deal.

Board question (Adam): if the ad is baked into the file and the file is fully downloaded, who exactly is the injured party — and is that an argument for the download metric or the final proof it's worthless?

Timing, for the cynics: Spotify's Q2 report credited "favorable podcast impacts" with pushing its profit margin to an all-time high — in the same window it built a button to strip other people's ads.

Pod News — Spotify threatens industry with ad-skipping tool

Semafor — Spotify's new 'skip' button could end the podcast business as we know it

Forbes — Spotify tests ad-skip button that could upend podcast advertising

Gizmodo — very cool for listeners, not for podcast businesspeople

Digital Music News — Spotify appears to be testing a 'Skip Ahead' button

-------------------------------------------------------------------------------------------------------------------------------------

Make the Funding Tag Buttom Prevelant!

-------------------------------------------------------------------------------------------------------------------------------------

01 - THE DEAFENING SILENCE — WHO ACTUALLY PUT THEIR HEAD UP

Cridland's charge, this week: "I find it fascinating, although not altogether unsurprising, that so many people who rely on podcast advertising for a living have sat on their hands and said bugger all on this." His read on why: nobody wants to go on the naughty step with Spotify or Apple and lose promotion.

Acast — Greg Glenday, given only after a week's thought (he declined to comment for the original story): "We do have some creators on the Acast platform who are not happy about this, and we will support them completely as they evaluate any recourse. Acast is working privately with Spotify to understand if we could have collaborated more on the rollout or positioning." Cridland's read: a hand-slap for not consulting partners, not a condemnation.

Adopter Media — Glenn Rubinstein calls it "fundamentally wrong." One of the very few naming it plainly.

Ben Richardson — interviewed on PNWR, sat squarely on the fence, neither for nor against.

πŸ‘€ Watch this space: Cridland said on air this week he is publishing another story next week — "because I've discovered even more." There is a second shoe.

Cridland on Spotify's character: "I would hope that Spotify respects the creators. All of the history of that company suggests that they don't." And on the replies he got: "the responses I've got from Spotify seem to show just a total misunderstanding of how the podcast industry works, which is more concerning to be honest."

The endgame he sketches: "The advertisers will notice when the ads don't work as well as they used to. And if the ads stop working, the advertisers will walk away. And we will have no opportunity to bring them back — they'll say, we've tried podcasting and it doesn't work for us anymore."

-------------------------------------------------------------------------------------------------------------------------------------

02 - THE APPS THAT GOT THERE FIRST — AD-STRIPPING AS A CATEGORY

The framing worth having: this is not one rogue app. In the space of about three weeks there are at least half a dozen shipping products whose entire pitch is removing the ads. Spotify didn't invent the behaviour — it industrialised it.

πŸ₯‡ PODTASTIC — the one Pod News covered (22 July 2026). Publisher: Freewheel Group Ltd. $2.99/month billed annually ($34.99/yr), or $9.99 monthly, 14-day trial. None of it goes to creators; up to 30% goes to Apple or Google. It "automatically skips past the interruptions other listeners skip" — crowd-sourced from aggregated listener behaviour rather than transcription. Plus AI summaries.

πŸ₯· Podtastic's real trick — it is nearly unblockable: the app downloads the MP3 straight to the phone and identifies itself in server logs as a web browser (Safari v604.1). To block it you'd have to block every web browser. The developer deliberately never mentions ad-skipping in the store listing and never names specific podcasts. Same developer previously ran PodcastAdBlock, which stripped ads and sold ad-free versions — shut down May 2026.

πŸ†• CASTRIA — the one Adam is thinking of, and it's FREE. iOS, iPad, Android, plus an installable web app, one account syncing across all of them. It transcribes the episode and pinpoints every ad, sponsor read and commercial break — including host-read ads, which crowd-sourcing misses — and skips them hands-free. One tap to switch it off or keep a sponsor you actually want. Ships AI summaries, searchable transcripts, tap-to-jump topics.

🎯 Castria's defence is the interesting part — they've pre-argued your side: "Does this hurt podcast creators? No. Castria downloads the complete episode straight from the publisher using your phone, so it registers as a genuine download and the ad impressions count just as the creator intended — we never re-host or alter the file." That is the download metric being used as a moral alibi. If the number is all anyone measures, stripping the ads is victimless by definition. Same trapdoor as Spotify's "we're not affecting ad delivery."

Also in the wild: ZeroAds (zeroads.ai) — "remove ads from any podcast, keep your app"; AdSkipPro; PurerPodcasts. There is now a comparison-shopping blog post ranking podcast apps by ad-skip quality.

Neil Vellio's voicemail to PNWR, which is the whole argument in ten seconds: "You can measure an ad to the end of its days, and it's still going to be an ad that someone's gonna pay to get shot of. Fix why people are paying to skip the ads, and half your measurement problem evaporates." And on the AMP Accords putting Spotify and YouTube on the committee: "a bit like inviting the two biggest meatheads to the anti-bullying committee."

Pod News — Another app promises to cut out the ads (Podtastic)

Castria

Podtastic

ZeroAds

-------------------------------------------------------------------------------------------------------------------------------------

04 - DAVE'S WORLD — PODPING v0.12.0, AND THE ONE-LINER OF THE WEEK

🎯 DAVE'S POST ON THIS EXACT STORY (7 Aug), the line to open with: "I look at what AMP wants podcast app developers to do in support of advertisers. Then I look at what podcast app developers are actually doing with advert skipping. One of those will win out. And it isn't the one that requires tons of first-party data collection/reporting and the taking on of privacy liabilities."

#podping v0.12.0 shipped (7 Aug) — writer, watcher and monitor, source and docker. DHT is gone: bootstrap peer discovery now uses the 5 stable, never-changing podping.cloud writer IDs as a pre-seeded list. Dave, 3 Aug: "It adds nothing but redundancy and bugs to the code."

The ask to the audience: "Existing watcher nodes will need to upgrade to v0.12.0 for long-term stability. I see hardly anyone has upgraded past v0.9.0. Upgrading is super easy now that everything is on docker hub." — make him say this on air; it's a direct call to action.

Iroh v0.97 → v1.0.3 across all podping.cloud servers (v0.11.0, 2 Aug). Multi-arch docker in progress from @suorcd; @DavidMarzalC testing on a Pi. Dave on a reported memory leak: QUIC connections historically eat memory per connection, so self-monitoring recycle code exists — "and build with --release, it makes a huge impact." He'll fire up an Iroh relay once the code settles.

πŸ’° The server-memory bit — good boardroom filler (13 Aug): "I just priced 128 gigs of ECC server memory from HPE. It's over $6,000." Then the correction: they don't sell a 128GB kit anymore, so it's two 64GB kits at $6k each — $12,000 for 128 gigs, as much as the entire base server itself.

Dave on the modern tech stack (12 Aug), replying to Marco Arment: "It's every corner of the modern technology corporate world. All the digital business processes are broken. Even the simplest things that used to be handled in 20 minutes can take days or weeks. All the beauty of modern tech interop gets stomped on by massive corporate greed, incompetence and inefficiency. When you make mountains of profit even while your shit is broken it's hard to get anyone to care." Also, flatly: "AI customer support chat bots are infuriatingly bad."

πŸ€– Gell-Mann amnesia, with a screenshot (7 Aug): "Gemini says Overcast supports video. Just your daily reminder to not believe what AI says without verification."

OpenJDK bans AI-generated contributions — "Contributions in the OpenJDK Community must not include content generated, in part or in full, by large language models, diffusion models, or similar deep-learning systems." Dave: "This coming from Oracle, who has bet their entire future on generative AI." His theory: Oracle's fear of copyrighted code surfacing in Java keeps the policy alive for years. Spurlock counters it reads as "interim" — "not no, but not yet." Dave: "Not yet can last a very long time."

Housekeeping: no board meeting on 7 Aug — Adam was out of town for the JCD memorial. Dead feeds are blocked naturally by the UNIQUE key on the url column of 'newsfeeds' — once marked dead it takes a new feed URL to get back in. And Mike Dell is speaking at Podcast Movement NYC: "Podcasting 2.0 — Why it's not Fringe anymore!"

Off-topic but very Dave (9 Aug), Wendell Berry: "Science can function as religion only by making two unscientific claims: that it will eventually know everything, and that it will eventually solve all human problems."

@dave — AMP vs. what app developers are actually doing

@dave — podping v0.12.0 ships, upgrade your watchers

@dave — Gemini says Overcast supports video

Show Notes

-------------------------------------------------------------------------------------------------------------------------------------

πŸ“Œ FROM PREP.md — Adam's PC20 items

Pod News — podcast app revenue (source for 'cut the apps into the money')

Pod News — the ad-skipping player

Board Q (Adam): "If you go to another chapter, is that a new play?" — for the AMP play=30s definition.

-------------------------------------------------------------------------------------------------------------------------------------

00 - DAVE'S WORLD: THE ALTERNATE ENCLOSURE FINALLY GETS PICKED UP + HLS-IN-RSS SURGE

Take the lap: Podbean and Buzzsprout are now testing/supporting the alternate enclosure — letting other apps (Pocket Casts, Amazon Music, iHeart, TrueFans) reach the video/alt versions instead of everything vanishing into the Apple-only void. Lead the boardroom with the namespace win.

HLS video in RSS is accelerating: @js — "the rate of podcast hosts publishing high-quality video in podcast feeds using HLS has gone up quite a bit, thanks mostly to Transistor and Flightcast." The open path is winning adoption; watch it live.

Namespace repo status: attribution bylines merged for the block and alternate enclosure tags (PRs #772/#773). Quiet week on commits — the plumbing is holding.

Dave in the wild: on the LLM-vs-tags thread — "A 'weight' is just a form of heuristic. Regression is literally built into the LLM at a fundamental level. Re-weighting a model will always regress other weights." (Continues the on-device-AI vs authored-tags fight from 265.) Also: Hover auto-renews domains 2 days out — @genebean's alert script gets a Dave thumbs-up.

HLS-in-RSS tracker (livewire)

Dave Winer — distributable is not distributed

-------------------------------------------------------------------------------------------------------------------------------------

BM - FROM THE TIMELINE — podcastindex.social this week (JS & Dave)

@js: "the rate of podcast hosts publishing high-quality video in podcast feeds using HLS has gone up quite a bit — thanks mostly to Transistor and Flightcast." (tracker: hls.livewire.io)

@js: "Waves of new podcast listeners may exist, but they aren't downloading episodes via RSS, no matter how many podcast apps are being vibe-coded."

@js shared Castro's "Podcast apps: a framework for dying" — his read: "well thought-out, but a bit too pessimistic imo — the next few years are going to be interesting." (Steven B: "anyone interested in the PC2.0 namespace read and consider this.")

@js on Patreon's "painful update" (Conte, layoffs): "I'd love to know what change Patreon's operational market has undergone in the last six months."

@js on the scraping wars: a court ruled aggregated search data (URLs, snippets, index facts) are "publicly accessible facts," not copyrightable — relevant to index/crawler rights.

@js (amplifying Dave Winer): "Threads and Bluesky are not distributed. Distributable is not the same as being distributed."

@dave: "A 'weight' is just a form of heuristic. Regression is literally built into the LLM at a fundamental level… one slider." — the on-device-abstraction vs authored-tags thread, continued from 265.

@dave, namespace repo: merged the block-tag + alternate-enclosure attribution PRs (mid-May); no new commits this week — the action is in adoption, not spec.

@csb → @dave @adam: "bros, I've just sent you $15.55 via bitcoin gram and amazingly only one split failed" — boosts still flow on-network (counter to 'streaming sats is over').

@mgdell (Mike Dell): "21 years of podcasting: the thing I don't want us to lose."

Mike Dell — 21 years of podcasting

Castro — Podcast apps: a framework for dying

-------------------------------------------------------------------------------------------------------------------------------------

01 - "WHAT IS A PODCAST?" — THE AMP ACCORDS PUT A FORMAL DEFINITION ON IT

The AMP Accords landed this week (Oxford Road's Dan Granger + Giles Martin), a year-long, commercial-end-of-the-industry group defining what a podcast IS, what a "play" is, and building exposure/attribution standards to replace the download. Note: the definition still names open RSS feeds.

The formal definition: "an on-demand show rooted in the spoken word and fully intelligible as audio… accessed as audio or video via open RSS feeds and other distribution methods." The commercial world catching up to what this room has said for years.

The play = 30 consecutive seconds (audio OR video); Spotify already adopting it, YouTube ~same ballpark. No time-window qualifier yet. Downloads called "a faith-based metric" (Tom Webster). "Amen," says James.

It's in the air everywhere: Mike Dell (@mgdell) — "21 years of podcasting: the thing I don't want us to lose." And @js flagged Castro's @dabluck must-read "Podcast apps: a framework for dying" (@StevenB: "anyone interested in the PC2.0 namespace should read it") — @js pushes back: too pessimistic, "the next few years are going to be interesting."

Mike Dell — 21 years of podcasting

Castro — Podcast apps: a framework for dying

-------------------------------------------------------------------------------------------------------------------------------------

06 - QUIPS / TRANSITIONS / LINK DUMP

-------------------------------------------------------------------------------------------------------------------------------------

Slop or Not! Lightning or Frightening! Adam and Dave go deep with the Boardroom after a 2 week break

ShowNotes

-------------------------------------------------------------------------------------------------------------------------------------

00 - PODPING GOSSIP-WRITER → PRODUCTION + INDEX PLUMBING

-------------------------------------------------------------------------------------------------------------------------------------

01 - THE OPEN INDEX TURNS UP INSIDE AN AI SCRAPER (SUNO)

@js flagged it: a hack at Suno reveals the AI music generator used PodcastIndex to do its targeting — "with the help of an online tool called PodcastIndex, Suno identified 420,000 different podcasts that had at least five 30-minute episodes and sought to download roughly 1 million hours of podcasts."

Board question: the Index exists to be "free, for any use" — and here "any use" is an AI company bulk-harvesting a million hours to train on. Is that the deal working as designed, or does anything change? A live test of the open-index ethos.

Ties straight to the on-device/namespace fight below and to the spam/AI-slop lane — the same extraction pressure from three directions.

Hack reveals Suno scraped via PodcastIndex (404 Media)

-------------------------------------------------------------------------------------------------------------------------------------

02 - ON-DEVICE AI vs NAMESPACE TAGS — "DIDN'T YOU INVENT THE NAMESPACE?"

Your own post set it off: Adam — Apple in talks to run a startup's compressed AI models on-device (PrisMML) — "More fodder for on-device features vs new tags." The core board fight of the week.

🎧 CLIPS TO PLAY — a 3-part PNWR sequence (Cridland and Sethi discussing your position; your own SOT is NOT in their show, so play + respond):

1. The jab (44.5s): recap of the Car Curious "podcast annotations" (≈ Apple timed links) debate — Nathan Gathright: LLM-from-transcript data stays the purview of players, not publishers; Sam Sethi: "Adam Curry also came in… he sees LLMs as the way for on-device abstraction of data rather than tags. And I'm like — didn't you come up with the namespace?" File: 01 - Cridland Sethi - Adam annotations, didnt you invent the namespace - PNWR.mp3

2. Sethi's thesis (55s): his actual position — fine, copy Apple, let LLMs do transcripts as a FALLBACK "if you can't be bothered," but if a creator wants to put real data in the feed, respect what the creator wants FIRST. "An LLM is a fancy autocomplete system… I'd much rather the creator is given the opportunity to do something good first." The publisher-vs-LLM argument, self-contained. File: 02 - Sethi - an LLM is a fancy autocomplete, creator first - PNWR.mp3

3. The Spurlock correction (36s): Sam apologizes for mischaracterizing @js — "I don't think we should leave it up to LLMs at all… but in practice it's going to be the app crawlers that mark shows up using hard work and drive for differentiation, not the publishers themselves." It might not be an AI LLM, but someone still has to mark the metadata up. File: 03 - Sethi - Spurlock correction, app crawlers not publishers - PNWR.mp3

Ammo for your side: @manton — his daughter's voicemail "Hi dad, this isn't an emergency" was transcribed by Apple as "this is an emergency." Player-side LLM inference is lossy; publisher-authored tags are authoritative. The case for the namespace in one anecdote.

Cultural backdrop (@js, Atlantic "The End of Reading"): a postliterate generation that treats reading the source as "arbitrarily withholding information" — if the phone abstracts everything for you, who ever touches the feed?

Apple + on-device AI compression (CNBC)

The End of Reading Is Here (Atlantic)

-------------------------------------------------------------------------------------------------------------------------------------

03 - FUNDING NAMESPACE: STREAMING SATS COLLAPSE → APPLE CASH IN podcast:funding?

Streaming sats fell off a cliff: PNWR — Sam Sethi says boosts/streaming sats are "down to pretty well zero" in July: only TrueFans and one from Podcast Guru, NOTHING from Fountain. "Streaming sats is going away quite fast now… a real shame." A direct V4V gut-check for the room.

They're quoting you: James Cridland cites "Adam Curry is talking about splitting the funding from the advertiser with the app or host developers" — app devs are otherwise "cut out of the entire economic chain." Your funding-split idea is now on the record on their show.

🎧 CLIPS TO PLAY — PNWR V4V audio (hosts discussing V4V; your SOT is not in their show, so play + respond):

1. Streaming sats collapse (47s): Sam Sethi — boosts/streaming sats "down to pretty well zero" in July; only TrueFans + one Podcast Guru, nothing from Fountain. "Going away quite fast now… a real shame. TrueFans is the place keeping it alive." File: 04 - Sethi - streaming sats down to near zero, TrueFans keeping it alive - PNWR.mp3

2. V4V economics (80s): James — app developers "weirdly cut off from the entire economic chain"; Sam cites your fix: "Adam Curry is talking about splitting the funding from the advertiser with the app or host developers… the ecosystem needs to have a way." File: 05 - Cridland Sethi - app devs cut out of the money chain, Adam funding-split - PNWR.mp3

Dave's concrete move: "Is an 'Apple Cash' request just a specially formatted IRI of some sort? I'm wondering if there's some way to do an open-ended AC request and plug that into the podcast:funding tag." A fiat on-ramp in the namespace as boosts fade — real board business.

Quip that fits here: Adam — "Try depositing $10k cash into someone else's bank account. They will surround you with armed guards within minutes." The banking-rails friction V4V was built to route around.

ADAM'S THOUGHTS

-------------------------------------------------------------------------------------------------------------------------------------

04 - CONTENT HIJACKING + AI-SLOP: SOL GOOD NETWORK AND THE SPAM-COP LANE

-------------------------------------------------------------------------------------------------------------------------------------

05 - MEASUREMENT: AMP'S "WHAT IS A PODCAST" PAPER DROPS NEXT THURSDAY

-------------------------------------------------------------------------------------------------------------------------------------

ShowNotes

-------------------------------------------------------------------------------------------------------------------------------------

-------------------------------------------------------------------------------------------------------------------------------------

Justin Jackson's post: We Have A Communication Problem

-------------------------------------------------------------------------------------------------------------------------------------

00 - PODPING.ALPHA STABLE + INDEX / PV-ALPHA HYGIENE

Milestone: Dave — "I think we can safely say that podping.alpha is stable now. It's been many weeks of 100% uptime." Lead the boardroom with the win.

PV-alpha 500s: @mitch + Dave debugging the "get a list of feeds that have updates over the last X" endpoint throwing 500s; @mitch adding a delay between paginated requests in case it's a too-many-requests block.

Feed de-listing puzzle: Dave to @ChadF — a feed marked dead with no spam flag; aggregators de-listing it for some other reason. Open question.

Iroh 1.0 — Dave flagged it: "Dialing keys instead of IP addresses." p2p networking, boardroom catnip — worth a riff with Dave.

Discussion: also surface your own snags — Sovereignfeeds webhook "Unknown Error sending to Server" and ladder.podcastindex.org appearing down (to @StevenB / Dave).

Iroh 1.0 Release

-------------------------------------------------------------------------------------------------------------------------------------

01 - SPAM + AI-SLOP — TOWARD A "SPAM-COP SCORE"

-------------------------------------------------------------------------------------------------------------------------------------

02 - MEASUREMENT: AMP'S 30-SECOND "PLAY," ROUND TWO

AMP defined a "play" as 30 seconds; Spotify immediately adopted it. You and Dave already called 30s "bullshit" last week — push total listen time + percent-completed as the real metrics.

The walkback: AMP's original press release said "30 seconds of content played… once per user per session" — and quietly DELETED "once per user per session," leaving a woolier definition. James flagged it on-page with the HTML5 del tag.

YouTube is in AMP: Google confirmed YouTube "has been participating in the AMP-led conversations" — implies YouTube uses a 30s play. Apple's stance still unknown.

RSS.com test (Alberto): moving the DOWNLOAD threshold 60s to 30s changed totals by ~1% — negligible. So 30s for both plays and downloads is just simpler.

The HLS gotcha: playing 10s of an HLS video podcast still downloads ~60-72s, so server logs can't tell real play — only player-side instrumentation (Spotify, future Apple) can. James: it's all too "cloak and dagger."

Watch July: AMP's implementation doc is due July and the group is light on technical people — needs to be real technical work, "not a sales press release."

Understanding podcast stats (PodNews)

-------------------------------------------------------------------------------------------------------------------------------------

03 - BEYOND CPM — INDIE + MISSION-DRIVEN ECONOMICS

Losh Moodaley ("Beyond CPM"): the indie middle class (5k-25k downloads = only ~1-1.5% of all podcasts) can't survive on CPMs. Roadmap: audience-as-economy, sell exclusivity not inventory, scale outcomes not audiences, "owners of trust, not renters of attention."

Pure V4V-adjacent framing — Sam Sethi tied it straight to TrueFans activity-based value (a share or comment is value, not just dollars) and "creator portals." Easy on-ramp to your worldview.

Vox Topica (Richard Fall): full-stack agency for nonprofits/mission-driven orgs — speaks engagement/reach/"depth of message," not downloads. Nonprofits resist AI voices (authenticity) but use AI for scripts/cleanup/show notes; now recommends video to ALL clients.

Beyond CPM: Surviving the New Measurement Era (PodNews Daily)

Vox Topica

-------------------------------------------------------------------------------------------------------------------------------------

04 - NUMBERS + MONEY MOVES

-------------------------------------------------------------------------------------------------------------------------------------

05 - AI BUBBLE / EDGE COMPUTE — DATACENTER WATER

-------------------------------------------------------------------------------------------------------------------------------------

06 - CROSS-STORY: COMMODORE FLIP-PHONE

-------------------------------------------------------------------------------------------------------------------------------------

07 - QUIPS / TRANSITIONS

-------------------------------------------------------------------------------------------------------------------------------------