Alberto Betella

Co-founder of RSS.com. Back in 2006, I developed Podcast Generator, one of the very first open-source web apps for podcast self-hosting, empowering a wide community of podcasters for over a decade.

Website: https://rss.com
Podcast Generator 2006 - 2026: https://podcastgenerator.net
LinkedIn: http://www.linkedin.com/in/albertobetella

Public Key

npub1tk4qp49vl72rl929z44v20y5ggkhjf8znv68z7z2ywqzz9yzxmzqux3wu7

NIP-05 Address

alberto_at_podcastindex.social@momostr.pink

Profile Code

nprofile1qqs9m2sq6jk0l9plj4z326k98j2yytteyn3fkdr30p9z8qppzjprd3qpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dsg9csym

Publishing to

offchain.pub

relay.damus.io

nos.lol

Author Public Key

npub1tk4qp49vl72rl929z44v20y5ggkhjf8znv68z7z2ywqzz9yzxmzqux3wu7

Show more details

Last Notes

2026-05-13 16:01:48 UTC

This cannabis-related podcast recommends 6,000+ podcasts via podroll 🤪 https://atlas.rss.io/#17222
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/568/088/605/041/527/original/7947f596cace214b.png

2026-05-13 11:46:31 UTC

- reply

David, Dave, it was a bug on my end, now fixed https://atlas.rss.io/#6571346
Initially the dataset only contained one item instead of arrays, then Dave improved it yesterday for me, and that left a bug in my code: https://github.com/albertobeta/podroll-atlas/commit/39fcdc13670ab82cecc8068142d16f0be88b7837

2026-05-13 11:37:19 UTC

- reply

M implementation simply fetches the JSON from PI's open dataset, caches it for 24h, adds a D3 layer on top of it, and has some UI/UX tweaks to make it more pleasant (e.g. "loading..." label when downloading metadata from external RSS feeds, fallback for CORS blocks, deeplinking to a specific podcast etc...). It's self-contained, and working fully in the client browser (no backend).

2026-05-12 20:08:37 UTC

- reply

Thanks. Some podcasts look beautiful when seen in the context of their network and community. e.g. https://atlas.rss.io/#540927
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/563/406/130/475/193/original/88f585ddda7397a0.png

2026-05-12 19:29:10 UTC

- reply

I think so. The repo with the source code is here https://github.com/albertobeta/podroll-atlas
I think that if you use this dataset https://public.podcastindex.org/recommendations_music.json
and AI and ask to adapt the code you quickly get a music version of my web viz

2026-05-12 17:39:08 UTC

- reply

yes, structurally this IS the podcast backlink. Also this is a real baseline for a search engine. The catch is coverage rather than the data model.

2026-05-12 15:07:11 UTC

- reply

I also added a custom domain so it's nicer: https://atlas.rss.io
e.g. No Agenda: https://atlas.rss.io/#41504

2026-05-12 14:52:50 UTC

- reply

Thanks Dave for shipping the full sources array in the dataset. Podroll Atlas now shows every podcast that recommends a given show (not just one example). Also added URL deep-linking via #feedId. e.g.
Podcasting 2.0: https://albertobeta.github.io/podroll-atlas/#920666
Podnews Weekly Review: https://albertobeta.github.io/podroll-atlas/#1333070 🚀🚀🚀
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/562/154/589/428/033/original/e00d7bce835ea895.png

2026-05-12 09:37:45 UTC

Thanks @npub1rcn…dw4h for releasing the open datasets from PodcastIndex: so much fun stuff you can build with them! e.g. I prototyped a Podroll Atlas, running entirely in your browser:
https://albertobeta.github.io/podroll-atlas/
Wishlist: if the recommendations dataset shipped the full array of sourceFeed IDs and URLs the graph would get a lot richer with clusters, fan-in, communities.
CC @npub1g2v…achf @npub13cg…ck4q @npub1thg…zq23 @npub1fmx…wp74
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/560/920/030/256/941/original/97ae38c3902ef0ac.jpeg
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/560/920/043/239/110/original/7929d7b0e8af3bc6.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/560/920/050/932/289/original/4180dad2da9a7c27.png

2026-05-07 18:06:44 UTC

Our first official video podcast: https://media.rss.com/podcast-network-insights/feed.xml
HLS via alternate enclosure, supported by default for every video episode on RSS.com.
Screenshot: same episode playing on Apple Podcasts and TrueFans.
@npub1g2v…achf @npub1thg…zq23 @npub13cg…ck4q
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/534/601/223/947/863/original/dbe5a811f111e7cf.png

2026-04-10 11:46:01 UTC

Closing the week on a personal note. This week 20 years ago, I released Podcast Generator, an open source tool I had built to help anyone publish a podcast. For the 20th anniversary, I dug through old screenshots, articles, and memorabilia and put it all together in one page. Took longer than I expected, but happy with the result: https://podcastgenerator.net
Nice to look back and remember those early podcasting days 💜
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/380/239/165/531/155/original/69f51d91352cd1b3.jpeg

2026-03-12 18:46:08 UTC

- reply

here is one RSS feed from Spreaker with the disclosure https://www.spreaker.com/show/6141933/episodes/feed
See also: https://validator.livewire.io/?validate=https://www.spreaker.com/show/6141933/episodes/feed

2026-03-12 18:41:22 UTC

- reply

I think they have to re-generate tens of thousands of RSS feeds and maybe they are doing that or perhaps they wait for the podcaster to update something in their show so the feed regenerates.

2026-03-12 15:10:59 UTC

- reply

Breaking just now: Spreaker also shipped AI disclosure 🤖 in RSS feeds using the TXT tag:
https://help.spreaker.com/en/articles/12458456-ai-generated-content-in-your-podcast-and-episode-descriptions

2026-03-09 15:04:23 UTC

- reply

Thanks. As a hosting our free plans do not allow DAI (it's a paid feature), so this is not about free tiers for us at least.
As a listener I am not concerned (eg I listen to finance AI podcasts), but I want to get to decide. Without a signal we don't have the choice. Same reason we label ingredients on food: most people don't read the label, but some do. And regulators (EU AI Act, FTC) are moving in that direction regardless. Better to have this in place now than to scramble later.

2026-03-09 11:05:14 UTC

Really enjoyed the latest P2.0 episode @npub1rcn…dw4h @npub1fmx…wp74
Just published a piece that also addresses some of the questions you raised https://rss.com/blog/ai-disclosure-in-podcasting-what-it-is-why-it-matters-and-how-to-do-it/
Excited about AI (we even hold a patent on AI voice cloning for ads). Ads in AI podcasts, not a big concern: if real humans listen, all good. But we like this voluntary signal to be there: it keeps podcasters ahead of regulators, and it allows apps to build filters or UI cues.
Also built a quick tool to help podcasters decide: shouldidisclose.ai

2026-03-05 15:16:50 UTC

- reply

with pleasure! Created PR 757

2026-03-04 21:35:05 UTC

- reply

I just built this https://shouldidisclose.ai by researching what every major platform and regulatory body requires and distilling it into something actionable today!
Rationale and framework of reference is here: https://shouldidisclose.ai/about.html

2026-03-03 17:11:06 UTC

- reply

wow! that’s really cool 🔝🔝🔝 Congrats Sam and TrueFans 🎉🎉

2026-03-03 12:16:59 UTC

- reply

Great news, Sam!

2026-03-03 12:14:35 UTC

- reply

We just published a short podcast episode about the challenge of AI in podcasting and the debate around it. It is fully written and voiced by AI based on public sources, and of course we properly disclose it.
We also used it to showcase a few 2.0 features: a Spanish audio version via alternate enclosure, chapters, a soundbite, location, etc. The RSS feed of that show uses 16 tags in total from the P2.0 namespace:
https://rss.com/podcasts/podcasting-innovation/2596546/

2026-03-03 05:53:03 UTC

- reply

Looks great! Love the robot icon

2026-03-02 13:45:24 UTC

- reply

Spreaker has been collecting this for a while (as a true/false value for now, like we do), and they add an automatic disclaimer in the episode description.
If Spreaker also added AI disclosure to their RSS feeds using the same tag and attribute we use, then in terms of episode market share (Livewire stats, Feb 2026) we would already be at around 17% of all podcast episodes produced in a given month including AI disclosure. That is almost 1 out of 5 new episodes.

2026-03-02 13:31:00 UTC

- reply

Nice! But there is no consensus yet on the scale and values being used. We started with a more generic top down approach. Spreaker is doing something similar, although they are not adding this data to their RSS feeds (at least not yet).
Glad to see we are moving toward AI disclosure. I prefer a top down approach cause it keeps the door open for future alignment. If everyone defines their own scales and values independently, it will be harder / unlikely to standardize later

2026-03-02 12:25:21 UTC

We just shipped AI disclosure to all our podcasters. No official framework exists in podcasting yet, so we used a deliberately generic checkbox.
Checked = true. Otherwise, false. Identical logic as Explicit.
We (think we) are the first to add this to RSS feeds. We chose podcast:txt (underused but versatile, and one of 3 P2.0 tags Apple recognizes) at the item level.
The sooner this signal lives in RSS feeds, the better (with NO penalties). We hope other hosts follow. https://validator.livewire.io/?validate=https://media.rss.com/podcasting-innovation/feed.xml
https://cdn.masto.host/podcastindexsocial/media_attachments/files/116/159/468/347/377/598/original/925401a4ca2b165b.png

2025-12-15 23:05:59 UTC

- reply

Nathan we did not move forward yet because we were trying to reach consensus, or at least a quorum. But those conversations on GitHub have had no interaction for the past 2 months.
Asking podcasters to tag when the host is AI, similar to how we ask about explicit language would not necessarily mean the content is slop (nobody would ever tag their episodes as slop), but it would allow more sophisticated behavior and better UX in podcast apps, and better data overall.

2025-12-15 22:43:12 UTC

- reply

This ties into the AI disclosure conversation,especially for AI hosts. I care less about AI used elsewhere in the workflow (eg cleanup),but asking to tag AI hosts would improve our data. There are 2 GitHub discussions on this,and I raised it in the PSP too,but nothing actionable came out yet. Spreaker is already asking for disclosure in their UI,but there is no tag to express it in RSS feeds.
We are not reacting fast enough. eg A new attribute for the person tag is a low-hanging fruit

2025-12-15 22:23:30 UTC

- reply

For the record: we did not release a public API to enable AI content dumping (if it happens, it is a byproduct). The goal is automations (Zapier, IFTTT) and integrations with third-party audio apps that have long asked to publish via API. And by “vibe coders experimenting with AI” I meant using AI to build automations faster. Zapier’s new “ask AI to build it” interface is just one example: I described a Zap that publishes on RSS.com via our API and it worked right away.

2025-12-12 11:15:51 UTC

For developers and “vibe coders” playing with AI: RSS.com public API is live 🤖
Can't wait to see the automated workflows and tools people build with this!
https://changelog.rss.com/v25-12-1-328320
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/706/305/940/264/394/original/b0f4173a18fbdd24.jpg

2025-10-27 13:15:41 UTC

- reply

This is exciting! If I could vote for a new regular dataset from Livewire, I’d love to see insights on location tag adoption: % of all episodes, % by host, % subject vs creator, and maybe a geo map (or a table like in OP3) with top countries or locations. In my "wildest dream", even the type of locations (e.g. “cathedrals”) though that requires cross-referencing the OSM API. This tag has huge potential. Tracking its MoM adoption would be fantastic. My two cents. cc @npub13cg…ck4q

2025-10-14 13:14:22 UTC

- reply

good examples! thanks

2025-10-13 18:20:22 UTC

@npub1g2v…achf the location tag 2.0 feels pretty novel and in the PSP we’ll be rolling it out to everyone, so adoption should grow fast. Would it make sense to keep a monthly stat on adoption? e.g. overall, by host, subject/creator ratio.
You already monitor most podcast RSS feeds globally, so pulling that data would be easier for Livewire than for others. We could do this for all the PI tags, but the location tag feels especially interesting for future discoverability: it could be a nice stat to share.

2025-10-02 05:33:32 UTC

- reply

Yes, that’s us! We have not changed the UA since adding Podping back in May 2021. We will update it this week to make it more recognizable.

2025-08-19 15:27:57 UTC

@npub1fmx…wp74 on stage at PM 2025. Go Podcasting!
https://cdn.masto.host/podcastindexsocial/media_attachments/files/115/056/120/874/956/923/original/fa0c77862e7df38e.png

2025-08-05 10:24:18 UTC

- reply

That said, I have always appreciated XSL's logic. Just last week I was testing a new XSL that we're about to roll into production: we had a feed with over 5k episodes and 2.5GB of data loaded each time it was opened in a browser, because the XSL rendered all episode artwork 🤯 I added logic to limit it to the latest 25 episodes. Pretty neat! Test feed: https://podcastgenerator.net/TESTS/newxsl/feed-large.xml
But honestly, I’d be happy with just the visual skin functionality.

2025-08-05 10:22:53 UTC

- reply

I just chimed in on the GitHub thread. For us, the main use case is adding a "skin" on top of XML: any lightweight solution that enables that would be enough.

2025-01-08 15:53:12 UTC

- reply

Our V4V UI was designed only for SSO with Alby for the sake of simplicity, but with the transition to Alby Hub and podcasters trying other solutions we are going to also allow to manually add a wallet. i.e in the case of your friend, you would disconnect from your Alby Wallet and manually insert your Alby Hub info. This is an improvement scheduled in our upcoming sprint (ETA end of Jan). I will ping you when released (great workaround with the 1% in the meanwhile! ☺️)
https://cdn.masto.host/podcastindexsocial/media_attachments/files/113/793/518/739/971/681/original/c144978fa9719257.png
https://cdn.masto.host/podcastindexsocial/media_attachments/files/113/793/518/926/874/675/original/905994a9b31af141.png

2024-11-30 18:49:16 UTC

- reply

No video podcasts for me, sticking to audio only 😉 The light is just for video calls.

2024-11-30 16:45:19 UTC

Upgraded my gear and put together a small podcast setup! New Year’s proposition: podcast more 🎙️✨
https://cdn.masto.host/podcastindexsocial/media_attachments/files/113/572/896/796/973/147/original/ce5ce6dd7c8f3340.png

2024-07-09 10:56:04 UTC

Question about Podcasting 2.0 Chapters endTime. If Chapter 1 ends at 1:00:00, should Chapter 2 start at the same time (1:00:00), at 1:00:01, or at 1:00:00.001? The specs only show examples without using endTime (https://podcasting2.org/podcast-namespace/tags/chapters), and the namespace definition mentions "The end time of the chapter, expressed in seconds with float precision for fractions of a second" [meaning? ms?] https://github.com/Podcastindex-org/podcast-namespace/blob/main/chapters/jsonChapters.md#the-chapter-object). Confused about using endTime and startTime for contiguous chapters. Thoughts?