Latest / WGKM / Podcasting 2.0 - Episode 269: Verbal Pollution

Transcript

Podcasting 2.0 - Episode 269: Verbal Pollution · WGKM

• 0:00
  Podcasting 2.0 for August 28th, 2026, episode 269, Verbal Pollution.
• 0:08
  Hello. Hello everybody, welcome to Podcasting 2.0. It's a Friday,
• 0:12
  it's 100 degrees outside. It's time for everything and all
• 0:16
  things podcasting. We're here to make it happen for you.
• 0:20
  It's the board meeting, and in fact, we are the
• 0:22
  only board meeting where we always blame Albie's. I'm Adam
• 0:26
  Curry here in the heart of the Texas Hill Country
• 0:28
  and in Alabama. The man who performed database surgery this
• 0:31
  weekend reports that the butt is intact. Say hello to
• 0:34
  my friend on the other end, the one, the only,
• 0:37
  the pod sage, Mr. Dave Jones. Everyone can... removed their
• 0:46
  hands from their butts. There's no reason to hold them
• 0:48
  anymore. It was the weirdest thing. The butts have landed.
• 0:53
  The butts have landed. You're like, the butt is okay.
• 0:55
  Like, all right. See, I saw that message before I
• 1:00
  saw, you know, the database message. Because it's a reverse.
• 1:03
  Oh, I bet that was confusing. Yes, the reverse chronological
• 1:06
  order for, Now I got to go figure out what's
• 1:08
  going on. Dave's in trouble. He needs help. We need
• 1:12
  to get to Alabama stat. He needs butt help. Butt
• 1:16
  help. How you doing, brother? Good. I'm good. I'm slightly
• 1:21
  grumpy from the... I have no chocolate, and that's really
• 1:25
  screwing up. No chocolate love. No, there's no chocolate love,
• 1:29
  and it's really messing with my head. Okay, so when
• 1:33
  you do... I have to imagine... That everyone is sort
• 1:39
  of like this who does podcasts. There has to be
• 1:43
  some sort of ceremony or ritual that goes along with
• 1:47
  doing each show. And I imagine that each show that
• 1:51
  you do is maybe a different ritual or ceremony, like
• 1:55
  pre-show, during show, this thing that you just have to
• 1:59
  like get. It sort of centers you into what the
• 2:02
  show is. Oh, that's an interesting question. Yes, I have
• 2:06
  different ceremonies for different shows. Yeah, chocolate is one of
• 2:11
  my ceremonies, one of my rituals, and it's really messing
• 2:15
  me up that it's not involved. I pray before every
• 2:18
  show. I do that. No agenda. I sing along with
• 2:22
  the fat lady. Yes. For podcasting 2.0, I play one
• 2:26
  of our Godcaster radio network stations and just listen along.
• 2:29
  Today was a little different as Tina is hosting a
• 2:33
  birthday party for one of her friends. And so there's
• 2:38
  eight women. right on the other side of the studio
• 2:41
  door. I'm sure you will hear them because they're church
• 2:45
  ladies. They get loud. Oh, yeah. And so part of
• 2:49
  my ritual today was shrimp, a boursin cheese with hot
• 2:56
  honey and pistachio on a cracker. You And a glass.
• 3:05
  On a cracker. Oh, man. Oh, that is so good.
• 3:09
  That is so good. Cheese, hot honey, and pistachio on
• 3:13
  a cracker. Well, no, it's Boursin cheese. You know Boursin?
• 3:16
  I don't know what that is. It's like the garlic
• 3:19
  spreadable French cheese. You've never had that? How do you
• 3:22
  spell it? B-O-U-R-S-I-N. Boursin. Now, most people say boursin cheese
• 3:29
  here in Texas. Boursin cheese. I'm like, ladies, it's boursin.
• 3:34
  Boursin. Yeah. You like the boursin, croissant, boursin, et le
• 3:39
  pain. Yes. And that's mixed with a little hot honey
• 3:45
  and then with the crushed pistachios as crack. The Europeans
• 3:50
  and their cheese, man. There's like 3,000 different varieties of
• 3:55
  cheese. Oh, yeah. And then the... In America, we have...
• 3:59
  American, which is not cheese at all. It's just slightly
• 4:04
  hardened milk. Plastic. Yeah, it's skim milk that's been curdled
• 4:08
  or something, and then we have cheddar. That's the glass
• 4:12
  of rosé I brought with me to the studio. Is
• 4:16
  this also a ceremony? Mm-hmm. It is today. It is
• 4:19
  today. Well, I'm good because you're grumpy and I'm not.
• 4:22
  So that's fantastic. I'm not I'm not super grumpy. I'm
• 4:25
  just mildly grumpy. Oh, OK. Boy, let's see that there
• 4:30
  are definitely some things going on. Yes. I think we
• 4:35
  should talk about the database surgery first. I mean, that's
• 4:38
  what I kick it all off with. So what was
• 4:40
  going on? I mean, your butt was hurting. The tables
• 4:43
  were large. There were outages. Linode was... was doing emergency
• 4:48
  migration. Why does that happen? Do they ever tell us
• 4:52
  why? Oh, emergency, emergency. We have to migrate your entire
• 4:56
  infrastructure to another server. Why? This used to never happen.
• 5:02
  Yeah. I mean, I've been using Linode since... I mean,
• 5:07
  since before we... Started podcasting, Podcast Index. So probably, I'm
• 5:13
  thinking probably make 2017, something like that. So probably a
• 5:17
  decade. It used to be very rare. Like once every...
• 5:22
  Like once every six to nine months that you'd get
• 5:24
  some sort of hiccup. And usually it was not that
• 5:28
  big of a deal. It was like some sort of
• 5:30
  like soft migration where they would, oh, this host has
• 5:33
  a disk degradation and we're going to migrate your VM
• 5:37
  to a new host. But we're putting you in the
• 5:39
  queue. It's going to happen sometime in the next 60
• 5:42
  days or 30 days. It was this like super chill.
• 5:47
  They weren't, you know, nobody ever seemed too concerned about
• 5:50
  anything. Ever since Akamai took them over, it's like this.
• 5:54
  They're going nuts over there. Things are dropping like flies.
• 5:57
  I just wonder what actually happens. I mean, it's a
• 6:00
  virtual machine, I presume. So how does that get corrupted?
• 6:04
  And there's never any, you don't really notice anything going
• 6:07
  wrong. They just tell you, well, it's emergency migration. I
• 6:11
  mean, the actual message was podcast index. No, that's not
• 6:16
  it. Where was it? Had it here. Well, this is
• 6:23
  from you. Linode Emergency Maintenance. That's usually what the message
• 6:28
  is. And does that come in on the Linode emails
• 6:33
  that you usually ignore, including Sam the showrunner, using all
• 6:36
  the bandwidth in the universe? No, I totally ignore the
• 6:40
  showrunner, but I never ignore the other, you know, the
• 6:45
  bat phone, essentially, which is like… Hey, emergency maintenance. Because
• 6:51
  this – well, so all this stuff is running on
• 6:55
  a rack of servers. And your host, your VM at
• 7:00
  any one time is on a server with – let's
• 7:05
  say 40 or 50 or 60 or 100 other VMs,
• 7:09
  depending on what the size of it is. And if
• 7:11
  that host hardware has some sort of problem, let's say
• 7:15
  usually, I don't know if they're using specialized hardware or
• 7:20
  vendor supplied hardware. I'm not sure. I don't know what.
• 7:23
  They're probably like super micro backed or something like that.
• 7:25
  But let's say, Let's say it's like a HP DL
• 7:28
  360. Yeah. Well, that thing has two power supplies. Well,
• 7:35
  if one of them dies. I forgot about that. I've
• 7:39
  had a power supply die on me, too, not too
• 7:41
  long ago. That's the weirdest thing. Yeah. It is because
• 7:45
  you don't expect it. That's so rare that it happens.
• 7:48
  But it seems like they're making the power supplies just
• 7:50
  powerful enough, you know. So that they don't have enough
• 7:54
  headroom and they're stressed. Yeah, because I have a Dell
• 7:57
  that I run everything on here. And maybe, was it
• 8:00
  maybe six weeks ago? All of a sudden, just boom,
• 8:04
  done. And I see the light flashing, and I go
• 8:08
  look it up. Oh, no, the light flashing means your
• 8:10
  power supply has failed. And then I look up the
• 8:13
  specs and everything. It's like, well, this thing is really
• 8:17
  kind of underpowered for if you max this machine out,
• 8:20
  which happens. Yeah, for sure. So I ordered a beefier
• 8:23
  power supply. I'm like, that kind of sucks. They give
• 8:27
  you just something that just kind of manages to get
• 8:30
  by. And this came off of an oil rig, this
• 8:34
  machine. One of our no agenda producers, he's a dude
• 8:38
  named Ben. He's on the oil rigs. He's like, so
• 8:41
  all these machines, all these beautiful small Dell machines, you
• 8:44
  know, like the little wedge. I forget what it is.
• 8:48
  Yeah. And there's just like hundreds of them doing stuff,
• 8:53
  I guess, for the oil rig. And they just have
• 8:56
  to rotate them out after so many hours just to
• 8:59
  make sure that the hardware is fresh. So I think
• 9:02
  they're probably all running at maybe 40%. of cpu so
• 9:06
  you know maybe that's how they deliver maybe that's how
• 9:09
  they ordered them just cheap jack man it makes sense
• 9:13
  though that a power supply would be a victim of
• 9:16
  a oil rig environment because you know that is not
• 9:19
  it's not that's what you call a hostile environment to
• 9:23
  electronics Yeah, kind of. Yeah, that would make sense that
• 9:30
  that thing would go kaput. Yeah. Out of any of
• 9:34
  the components. Well, luckily. It's sucking in air all the
• 9:36
  time. It's got a cycle in that sea air through
• 9:39
  it all the time. Well, luckily, he sent me like
• 9:42
  five of these, so I had another power supply. I'm
• 9:44
  like, well, it's the same power supply. I'll order a
• 9:47
  bigger one. Oh, yes. Yeah. Yeah. Yeah. So anyway, so
• 9:53
  everything came back okay. But then you were doing other
• 9:55
  stuff. You were doing something with a big table or
• 9:58
  something.
• 10:03
  So I'm really glad that this thing, that the Linode
• 10:06
  maintenance did not happen at the exact moment that I
• 10:10
  was trying to do this other stuff because that would
• 10:12
  have been really bad. We talked about this a while
• 10:17
  back briefly. about what this thing is, and it's just
• 10:22
  taken me a while to get around to actually pulling
• 10:25
  the trigger because these things take a long time, and
• 10:28
  it's still sort of in a rolling plan mode. I
• 10:33
  won't say that I have this master runbook that I'm
• 10:36
  following. This is really just sort of I'm feeling my
• 10:40
  way through this, End. But it requires such a big
• 10:46
  refactor that that's one of the reasons I was kicking
• 10:50
  the can down the road is because now I feel
• 10:54
  like I have enough agentic help sort of to help
• 11:01
  me review stuff. to be a reviewer for my for
• 11:06
  my you know how i think this thing is going
• 11:09
  to go like because okay so if you look at
• 11:12
  the where did i post that let's see is it
• 11:17
  api podcast index Or, um, I posted like an open
• 11:24
  source version of the podcast index API core, one of
• 11:28
  the low level include files a while back, uh, github.
• 11:35
  I'm sure I can find it. Here it is. The
• 11:38
  big schema refactor. Is that it? I see. Where are
• 11:42
  you seeing that? No, I'm just, that's wrong. That's not
• 11:45
  it. Docs API. Thank you. Let me scroll down. I'm
• 11:52
  trying. I wanted to be able to drop something in
• 11:55
  so that people could look at why this, okay, maybe
• 11:58
  this is API v1. Yeah. Okay. All right. So here's
• 12:04
  a good example of why this is complicated. Where's the
• 12:10
  boardroom? There you are, boardroom. Hello, boardroom. Hello. This is
• 12:16
  the open source... version, so to speak, of the of
• 12:24
  one of the core files there's a bunch of these
• 12:26
  now when you when you say open source so this
• 12:29
  is not the full production code this is just something
• 12:32
  you're giving to saps who think they can emulate what
• 12:35
  you're doing is that what that is no way no
• 12:38
  no this is this is uh So this this is
• 12:41
  something I cleaned up and because to me, this stuff
• 12:44
  is useless without comments. Right. And so I had to
• 12:48
  go through and just like comment all the code and
• 12:51
  try to get it just in a in a form
• 12:53
  that's usable where people could even see what the heck
• 12:56
  was going on. So this is I mean, this is
• 12:58
  the actual code. But I just cleaned it up a
• 13:01
  lot to give it a lot of... Because these comments
• 13:03
  and stuff were not there. And so this... Is really
• 13:10
  how big, so this is just one, this is 14,000
• 13:15
  lines, I don't know how many functions, but these are
• 13:18
  sort of the low-level functions of the API where they
• 13:21
  communicate directly with the database. So most all of these
• 13:25
  are just database calls. Mm-hmm. So when you look, when
• 13:29
  you do like a podcast by feed ID, it's going
• 13:32
  to call these functions on the backside to get the
• 13:35
  data out of the database. Well. So this refactor, let
• 13:42
  me back up. So this refactor, The way that the
• 13:47
  podcast index database came about initially – Because we sucked
• 13:55
  everything out of Apple. Shit. You just sucked it back
• 14:03
  You gave them iPodder so we just sucked it back
• 14:06
  We did a back suck It was a fair swap
• 14:09
  Yeah it was just a back suck Back suck Yeah
• 14:14
  Just move on from that place Just move on But...
• 14:22
  Where do I'm translating? Where to start? Okay, well, the
• 14:26
  origin of the database schema originally comes from Freedom Controller
• 14:30
  because that was RSS Ingestor, Ingestor, Ingestor, Ingestor, whatever. So
• 14:37
  the name of the database that held all the feeds
• 14:42
  was… News feeds or the table that had all the
• 14:45
  feeds was news feeds. And then you had NF items,
• 14:48
  so news feed items was where all the items went.
• 14:52
  So that became NF. Those things became synonymous with podcasts
• 14:57
  and episodes, yes. So news feeds just became an alias
• 15:02
  for podcasts. Right. Um, Because... My thinking forever has always
• 15:11
  been that the... That a podcast is a, is a
• 15:15
  feed. I mean, because, because it is, it is, but
• 15:19
  it's not. Oh, yeah. So the podcast both is and
• 15:27
  is not a feed at the same time. So the
• 15:30
  podcast. The feed defines the podcast. but... But the podcast
• 15:41
  itself is also represented by the feed. Hmm. This is
• 15:48
  a cosmic mystery that we can't fully grasp. It's like
• 15:55
  the Heisenbergian podcast principle that you're both – That you're
• 16:02
  both a feed and not a feed at the same
• 16:04
  time. No, Schrodinger. This is Schrodinger's podcasting. But if you
• 16:09
  think about it, because if you think about it in
• 16:11
  one way, the podcast doesn't exist to the world. Unless
• 16:17
  there's a feed. That conveys it. But at the same
• 16:23
  time... A podcast is not simply the feed. No agenda
• 16:29
  is a podcast regardless of – even if you deleted
• 16:32
  the feed today, no agenda would still be – In
• 16:36
  One Sense, a podcast. The episodes would still exist somewhere.
• 16:43
  You could send the No Agenda podcast to somebody tomorrow
• 16:48
  if the feed was deleted. You could send them all
• 16:50
  as a zip file, and they could listen to all
• 16:51
  of them in sequence. True. So, I think, In some
• 16:58
  ways, this solves a lot of problems that we've had.
• 17:02
  And everybody will be familiar with these problems. Yeah. One
• 17:08
  of the problems is duplicate feeds or conflict feeds. Somebody
• 17:15
  moves their podcast from one host to another, and now
• 17:19
  you've got two feeds. Well, which one is right? How
• 17:22
  do we know which one is right? They did it.
• 17:24
  Maybe they put in a redirect from the old host
• 17:26
  to the new host. Maybe they didn't. Um, We have
• 17:31
  no control of it and no real way to know
• 17:35
  which one is good. Canonical. So, What we do is
• 17:43
  this convoluted dance of trying to figure out through a
• 17:47
  bunch of heuristics which – which one to treat as
• 17:51
  the authentic feed for this podcast. Right, which is really
• 17:56
  the trick of the whole index is to be able
• 17:58
  to do that. Yeah, and so we do that primarily
• 18:03
  through this function. The news feeds table has a couple
• 18:09
  of columns in it. One of them is dead is
• 18:16
  the name of the column, and it's just either a
• 18:19
  one or a zero, and the other one is duplicate
• 18:22
  of. which takes a podcast index feed ID. And so
• 18:28
  you can – the way we would do that is
• 18:31
  no agenda moves from its current feed URL to a
• 18:34
  new feed URL tomorrow. We mark 40. So then now
• 18:41
  so now and the podcast index picks up this new
• 18:44
  feed. So now you get two podcasts IDs in the
• 18:47
  news feeds table. They both say no agenda. And one
• 18:53
  of them is 41.504, and the new one is 8
• 18:56
  million-something-something. Mm-hmm. The way we would properly treat this currently
• 19:05
  today is we would say – We would... We would
• 19:11
  mark the 8 million and something, the new one, we
• 19:14
  would mark it as dead. change its URL because in
• 19:19
  the news feeds table, the URLs are unique index. So
• 19:22
  they can't, you can't have to do the same one
• 19:24
  on different rows. We would mark it. We would change
• 19:29
  the URL to something like the URL underscore conflict just
• 19:35
  to make it be something different. Then change 41504 to
• 19:41
  be the new URL that the 8 million something used,
• 19:45
  you know, did have. And then in the 8 million
• 19:48
  something feed ID row, we would mark, we would set
• 19:53
  duplicate of to 41504. So now you have sort of
• 19:58
  an audit trail of...
• 20:01
  These are all the feeds that are duplicates of the
• 20:04
  original feed, but we want to maintain a canonicity of
• 20:08
  the 41.504, so we always – blow away these new
• 20:14
  versions and keep replacing 41.504's feed with whatever the current
• 20:19
  one is, as best we can determine. Is Melissa still
• 20:21
  living in the same house as you? Laughter school started
• 20:27
  back so she's teaching so she does not have to
• 20:30
  deal with me right now okay okay just checking just
• 20:32
  checking this is i'm getting the buzz again are you
• 20:36
  hearing this buzz yeah i hear the buzz it's no
• 20:39
  chuckle love man it went away yeah So that's how
• 20:44
  we've always done it. And it's fine, but it is
• 20:48
  a big dance to go through. Yeah. And it also
• 20:51
  sometimes presents problems. that are sort of hard to get
• 20:58
  yourself out of. Because you can imagine, that was a
• 21:02
  very simplistic example. Please, don't give me the complicated one.
• 21:07
  Please. Some podcasters... are nuts. And they go through seven
• 21:15
  hosts before they figure out where they want to be.
• 21:18
  I know. How do you get out of that mess?
• 21:22
  Um... So – Here's And Daniel says, I think you're
• 21:29
  trying to take a burden on your shoulders that should
• 21:31
  not be yours to bear. You're right. I'm bearing a
• 21:39
  burden. But – So it struck me – This notion
• 21:49
  that the podcast is not the feed. And so what
• 21:58
  we really need... Is a layer above, an identity layer
• 22:05
  above the feeds table that is a podcast table. Wow.
• 22:14
  Okay. Is this blowing you away a little bit? Well,
• 22:23
  I mean, the minute you say that the podcast is
• 22:26
  not the feed, I mean, yeah, in the database, I
• 22:28
  understand because there can be multiple feeds and it's a
• 22:34
  mess. But let me back up a little bit. There
• 22:40
  has been an overall failing in the podcast industry, if
• 22:45
  I may call it that for a moment, because I
• 22:48
  think it is partially the podcast industrial complex is to
• 22:51
  blame. Because I talk to young people. I talk to
• 22:56
  the young'uns, my nephews mainly. And they're like, yeah, we're
• 22:59
  starting a podcast. What are we going to do? I'm
• 23:01
  going to put it on Spotify. I'm going to put
• 23:03
  it on Spotify. What does that mean? Well, we're going
• 23:05
  to, you know, we've got free stuff. We'll put it
• 23:08
  on Spotify. How about Apple? Yeah, we don't think we
• 23:13
  want to use any other, you know, it's like, how
• 23:17
  do we do that? I said, okay, did you read
• 23:20
  anything about it? No, no, I just want to be
• 23:24
  on Spotify. Well, I can set you up with an
• 23:27
  RSS.com account. Oh, that's one of those distributors. This is
• 23:33
  a distributor. Yes, this is a critical, critical piece of
• 23:37
  information that I got. Because of the lack of, well,
• 23:42
  it's not just the lack of education, but it is
• 23:47
  the marketing of YouTube and Spotify and others that has
• 23:55
  made newcomers think that, Podcast hosting companies are distributors. And
• 24:05
  honestly, if you look at RSS.com, which I use for
• 24:08
  a couple of podcasts, In fact, I'll tell you exactly
• 24:13
  what it says. It's probably worth it. Let me take
• 24:16
  a look. So if you go rss.com, and look at
• 24:21
  my – go to my dashboard. And it says right
• 24:25
  there at the top, I've got my podcast episodes analytics
• 24:29
  distribution. With a little share icon, you know, the little
• 24:33
  three dots, one dot to the left and two dots
• 24:37
  connected by wires. And it says major directories. And it
• 24:42
  has all these different switches. So they are, in effect,
• 24:47
  teaching me as someone who knows nothing. That they are
• 24:53
  a distributor because now in the distribution tag tab. So
• 24:58
  this is why so many people say you haven't uploaded.
• 25:03
  You forgot to upload to Apple. No, I didn't really.
• 25:07
  So actually I did not. But this is a huge
• 25:10
  disservice to themselves. Because the whole, and also people saying,
• 25:16
  open podcasting. You hate this term. I do. Now, some
• 25:25
  might say open RSS. Okay, I can kind of live
• 25:29
  with that. But there is no more knowledge of how
• 25:34
  it works and why it works that way and what
• 25:36
  the entire concept was and what the benefit is to
• 25:41
  having an RSS feed. And to be fair about it,
• 25:45
  people, well, I got a podcast. It's on YouTube. I
• 25:48
  put it on X. And, of course, with the disastrous
• 25:53
  results when X or YouTube or whatever takes your podcast
• 25:58
  off, it's like, no, you're done. So – This is
• 26:02
  kind of interesting that you then say to me, the
• 26:05
  podcast is not the feed. For the same reason, you
• 26:08
  have podcasters who are, quote, nuts, because they're just trying
• 26:13
  different distributors, man. Yeah, they're looking for the cheapest one.
• 26:17
  The hosting companies themselves, I believe, are partially to blame
• 26:21
  for this. There's no educational – I don't see the
• 26:25
  process at rss.com where – and it's in the name,
• 26:29
  rss.com, explaining why. It just says distribution. Distribution. Now, from
• 26:36
  a... flow perspective, I kind of get it, but this
• 26:42
  is the downstream effects of the lack of education, of
• 26:46
  the sheer beauty of RSS, certainly when it comes to
• 26:52
  podcasts, the sheer beauty and the whole beauty of the
• 26:56
  system is, Having a distribution system that you control from
• 27:02
  an RSS feed with your files attached to it, that
• 27:06
  is, I'd say, pretty much lost at this point. And
• 27:09
  here's the downstream effects of it. You're saying it's lost
• 27:14
  in the sense that no normie understands that process. Worse,
• 27:20
  they believe it works a completely different way. Right. Yeah,
• 27:26
  so the podcast, well, that's a good example, your nephews.
• 27:34
  of The idea that they probably think that they probably
• 27:39
  think that it operates a different way because the podcast
• 27:44
  is a priority to the feed. They in their mind,
• 27:49
  they already have a podcast. It's got a name. It's
• 27:53
  got a topic. It's got the hosts. They probably –
• 27:57
  maybe they already have some album art or something like
• 28:02
  that, and they may already have the first episode in
• 28:06
  the can. But they just don't have it on Spotify
• 28:09
  yet. So the podcast to them already exists. But they
• 28:14
  just have to distribute it. They got to get it
• 28:17
  out there. But we take the sense that if the
• 28:20
  feed doesn't exist, the podcast itself doesn't exist. Which is
• 28:25
  not entirely – which goes against what the normie believes.
• 28:30
  Yeah, but I've always hated the term normie, to be
• 28:32
  honest about it. Well, I'm just shorthanding that. Yeah, okay.
• 28:38
  To differentiate between – I mean, we're doing a –
• 28:43
  It was widely understood. You need an RSS feed. That
• 28:49
  was it. And I'm talking back to 2015, 2016. Widely
• 28:57
  understood. You need an RSS feed. It is no longer
• 29:00
  understood. And that's part of the whole problem, the part
• 29:02
  of the overall issue that we're seeing just across the
• 29:06
  board with everything. That's why people go to YouTube. That's
• 29:10
  why people don't switch on their RSS feed when they
• 29:13
  use Spotify for creators. Eric PP says, I hate Normie
• 29:18
  too, is demeaning. Let's dive in on this for just
• 29:22
  a second. Normie, you can take – I had a
• 29:27
  conversation this week about this very topic. is. The average
• 29:35
  person... In corporate America. Misunderstands the emphasis of the term
• 29:43
  artificial intelligence. You can take this thing in two different
• 29:47
  ways. You can say you can mentally place the emphasis
• 29:52
  on the term artificial intelligence. So in your mind, you
• 29:57
  can see, and the more you use
• 30:00
  this stuff on a technical level, the more you do
• 30:02
  place the emphasis, you can say, I'm using artificial intelligence.
• 30:10
  Meaning, in your mind, you see the artificial part, this
• 30:14
  thing as the overriding factor. This is... It's an intelligence
• 30:20
  which is artificial. It's not genuine, so therefore I distrust
• 30:25
  it. Or maybe not just distrust it, but I treat
• 30:29
  it with everything it says with a grain of salt.
• 30:34
  The average corporate America user that is not a technical
• 30:37
  person, meaning they are not steeped in computer technology and
• 30:42
  the day-to-day imaginations of the technology industry, tend to place
• 30:47
  the emphasis on the second word. And they, they, in
• 30:52
  their mind, it is artificial intelligence. Okay, stop. Go back
• 30:55
  to the beginning. It comes from the technology industry who
• 30:59
  is treating users as normies from the beginning. This is
• 31:03
  artificial intelligence. Fully well-knowing, it's not. And then, but the
• 31:10
  normies will think this, that is the arrogance of the
• 31:13
  technology industry across the board. You're holding it wrong. Absolutely.
• 31:20
  I think you're exactly right. I would not disagree with
• 31:24
  that. When it comes to the term normie, though, You
• 31:30
  can use that term as a person. Yeah, but it's
• 31:33
  like premium. But Eric said it's demeaning. Yeah, I think
• 31:38
  it is. I agree. It's only demeaning if you say
• 31:42
  – if you come at it mentally from saying I'm
• 31:46
  the technical – technically superior person and you're just a
• 31:51
  normie or the way I meant it was I'm the
• 31:57
  weird one, and you're normal. I'm somewhat of an oddity
• 32:02
  in the idea that I understand the XML of an
• 32:07
  RSS feed, and I think that's honestly a little strange.
• 32:13
  I don't think that a normal person should have to
• 32:16
  understand this. It's not a demeaning term. It's a term
• 32:21
  of attribution of a certain set of qualities that I
• 32:26
  think is actually normal. Okay, I submit this to you.
• 32:31
  Because of the exact thinking you just displayed, which is
• 32:35
  not demeaning, it's not demeaning. Because you're a good guy.
• 32:41
  You don't mean any. You never mean any. I know
• 32:43
  you. I've heard you in arguments with people and you're
• 32:46
  just still a loving guy. I've heard you snap, though.
• 32:51
  That was kind of fun. But anyways, there is a
• 32:53
  snapping point. But because of this segregation, I believe that's
• 33:00
  exactly why we wound up in this spot. Well, we've
• 33:03
  got to make it easy for normies to understand. Just
• 33:08
  call a distribution. I think that's where the problems stem
• 33:12
  from. So not even from a degrading standpoint, but we're
• 33:18
  the people who get it and those are the people
• 33:21
  who need to use it. They'll never get it. So
• 33:23
  let's just simplify it without thinking about the words that
• 33:26
  we're using because that'll just work for them. And this
• 33:30
  is the downstream effect of it. So not even intentionally
• 33:36
  demeaning In fact, probably quite unintentional, but that's what happens.
• 33:44
  I mean isn't the whole thing the result of just
• 33:48
  – I think your original idea of this is all
• 33:53
  downstream from Silicon Valley is probably the truth. Yeah. Nobody
• 33:59
  – so let's go back to – The 19 –
• 34:05
  let's go back to the 1920s. Go back to the
• 34:09
  Model A and the Model T Ford. When you drove
• 34:15
  this thing, it wasn't – if you were going to
• 34:18
  – if you bought a Model T – And you
• 34:21
  drove it on a daily basis. The amount of things
• 34:25
  you had to go through. But there was a –
• 34:28
  so the internal combustion engine – It has to have
• 34:34
  ignition within an internal combustion engine. So as the piston
• 34:39
  approaches the top of its travel. Maximum compression. Maximum compression
• 34:46
  of the air and gas mixture. And the moment that
• 34:52
  the spark plug fires to ignite it, which will then
• 34:56
  drive it down, And spin the crankshaft on the next
• 35:00
  phase of its rotation. Yes. That point of ignition... The
• 35:07
  explosion, an ignition of an explosive material is a static
• 35:15
  event. It happens over a static amount of time. But
• 35:20
  as the engine goes faster, the time that is allotted
• 35:26
  for that to happen is... gets shorter and shorter and
• 35:30
  shorter. So the explosion inside of a piston is a
• 35:33
  static event inside of a dynamic environment. So in order
• 35:38
  to account for that, you have to start the ignition,
• 35:42
  that point of ignition within the cylinder. You have to
• 35:45
  start that process sooner. It has to happen earlier in
• 35:50
  order for the explosion to finish at the correct moment.
• 35:58
  where the piston is going to be in its travel.
• 36:01
  So there's a thing called ignition advance. So as the
• 36:05
  engine goes faster, the spark has to start earlier and
• 36:08
  earlier and earlier. Okay. I'm familiar with the workings of
• 36:14
  the combustion engine. You are. You are. But we get
• 36:18
  nerds in the boardroom that don't know. No, no, no.
• 36:22
  Stop. This used to be taught. In schools. Shop class.
• 36:28
  Yes. There is no shop class anymore. And exactly my
• 36:32
  point. There is no education. But we as a podcast
• 36:37
  community have failed with the education and the... Commercials. That
• 36:52
  is not acceptable. That's not acceptable. The monies, people who
• 36:58
  care about monies... The product. Yeah. They have... They hijacked
• 37:06
  this to the point now where they're being eaten themselves
• 37:11
  by nonsense that has nothing to do with podcasting, such
• 37:15
  as YouTube, etc. Because, And, you know, I'm sure you
• 37:21
  saw the story about the children's channel that was taken
• 37:25
  down from YouTube. And, you know, we don't know why.
• 37:29
  But thank God, thanks to the 33,000 readers of James's
• 37:32
  email newsletter, the YouTube caught wind of it and brought
• 37:36
  it back. But the whole thing should have never happened
• 37:39
  in the first place. It really shouldn't have. That was
• 37:47
  the whole point of it. Mm-hmm. So anyway, we don't.
• 37:53
  Yeah. I know you mean well. Well, I mean, I
• 37:57
  think you're on the right track because the point I
• 38:02
  was trying to make with all that is that in
• 38:05
  the early days of automobiles, this process of advancing the
• 38:10
  ignition the faster you went was a stick on the
• 38:12
  side of your steering wheel. Yeah. Yeah. I know you
• 38:16
  had to manually click as you went faster. The engine
• 38:19
  would start to drive like crap and you would click
• 38:22
  it down to the next setting. And then you went
• 38:24
  a little bit faster and you click it down to
• 38:25
  the next setting. This was a manual thing you had
• 38:28
  to do as you drove. Okay, so now what you're
• 38:30
  talking about. And everybody knew this. Right. And my dad
• 38:33
  had Model A's. And he and his brother would crash
• 38:36
  them and stuff. So I've heard plenty of stories about
• 38:39
  Model A's. But... So now we get into basically the
• 38:43
  crux of it is because there is no education of
• 38:46
  how things – whenever someone says, hey, I want my
• 38:49
  kid to learn about computers, I say, toss him an
• 38:53
  old laptop, give him a USB stick with Ubuntu on
• 38:56
  it. Tell them to install it and figure it out
• 38:59
  and tell them to come back when the thing is
• 39:01
  working. It's on Wi-Fi and they've learned some stuff. This
• 39:06
  is the same when I, you know, I've been flying
• 39:08
  since 2006. I wanted to be able to rent a
• 39:13
  plane, so I go to the... to the airport in
• 39:16
  Bernie, and I need to be educated on the Cirrus
• 39:21
  aircraft platform, which is all garmin g1000 i've never flown
• 39:25
  with it or i had never flown with a glass
• 39:28
  cockpit i use the six pack six dials you know
• 39:31
  you got your you got your airspeed indicator got your
• 39:34
  horizon uh you've got your attitude indicator and that's really
• 39:38
  it you don't need much more but you got you
• 39:41
  know some engine management so i'm in this uh in
• 39:45
  the Cirrus, and we're doing an instrument approach, and the
• 39:49
  instructor says, okay, you know, so do your teardrop now,
• 39:54
  you know, this particular method you take. And I've done
• 39:58
  these before, it's just I'd never done it on.
• 40:00
  computerized airplane. And so the way you typically do it
• 40:05
  is you have a standard rate turn and you have
• 40:08
  a stopwatch that you used to just literally have a
• 40:10
  stopwatch in the cockpit and you hit it, you know,
• 40:13
  after 30 seconds, you know that you're exactly back at
• 40:15
  the right end of the teardrop. Sometimes it's a minute.
• 40:18
  It depends. But, It depends on the chart and the
• 40:21
  approach plate and everything that you have. And I look
• 40:23
  at this cockpit and it has arrows indicating everything that
• 40:27
  I have to do. And I say, wait a minute,
• 40:30
  can you use this in the test? They say, oh,
• 40:33
  yeah, yeah, yeah. When you do the test, you just
• 40:36
  fly the arrow. which is great but if that fails
• 40:43
  people will die because they don't know how to do
• 40:46
  it anymore so that is in general where technology always
• 40:50
  takes us and so now we're you know what this
• 40:53
  is this is called techno solutionism oh i like that
• 40:56
  that's a good phrase I heard that this morning. This
• 40:59
  is TechnoSolutionism. Yeah, well... It's – some technology company somewhere
• 41:09
  gives you a solution to a problem, and it ratchets
• 41:13
  the knowledge base of the population down a notch, and
• 41:17
  you can't – and now you're completely dependent on the
• 41:22
  thing. And it's not like that's always a flat negative,
• 41:30
  but it's the same reason that if ever I was
• 41:36
  learning a new programming language or something like that, I
• 41:40
  would always go back – And by the – I
• 41:44
  would try to find the original books of whatever that
• 41:49
  thing is, like C. You go back and you get
• 41:52
  a copy of the C programming language by Kernighan and
• 41:56
  Ritchie or Java. You would get the original Java documentation
• 42:02
  from Sun Microsystems. But you go try to find the
• 42:05
  earliest incantation of that thing so that you could, like,
• 42:10
  get familiar with what the base level of the technology
• 42:14
  was and then move your understanding forward through history to
• 42:17
  try to catch up. But you had to make sure
• 42:19
  you knew the basics first because otherwise you're just at
• 42:23
  the – you're always being – Driven by the whim
• 42:26
  of whatever the latest technological feature at, you know, Sprint
• 42:31
  is or whatever that you never really fully get an
• 42:35
  understanding of of the foundations of why the stuff is
• 42:38
  the way it is to begin with. And I mean,
• 42:43
  maybe to me, that's what happened. I mean, it's podcasting.
• 42:48
  I mean, in the original days, everybody understood that the
• 42:52
  podcast feed – Because you had to copy and paste
• 42:55
  that thing into your software. Yes, into your podcatcher. Yeah,
• 43:00
  but then it went into iTunes, and the whole idea
• 43:05
  that there even was a feed was just disappeared. Well,
• 43:11
  Not from the hosting side. That's where the idea disappeared.
• 43:18
  I'm kind of blaming the hosters. Well, I mean, that's
• 43:23
  a solution to sort of a problem, I guess. Anyway,
• 43:28
  so take me back to the index. So now the
• 43:31
  feed is no longer the podcast. So what is it?
• 43:39
  The feed is what happened to the feed is the
• 43:41
  source of truth. It still is. It still is because
• 43:45
  the feed is how the podcast works. you know, represents
• 43:50
  itself to the world. But now there's the – it's
• 43:54
  just no longer the identity layer, the top level of
• 43:58
  the identity layers. So you have a podcast. So there's
• 44:04
  – as the database stands – Today, as we speak.
• 44:10
  The podcast table exists in the index database. And it
• 44:16
  has all the proper... All the proper metadata columns. So
• 44:24
  let me pull it up. Let's see. Let me refresh
• 44:30
  this. Okay. So there's now a podcast table. that has
• 44:39
  A feed ID, excuse me, has a feed, an ID,
• 44:43
  a feed ID. An iTunes ID. The GUID. and a
• 44:52
  few other incidentals. So dead being one of them. So
• 44:57
  now the next thing that's going to happen is there's
• 45:01
  going to be a seeding process. where The news feeds
• 45:09
  table is going to begin to populate The podcast table
• 45:15
  with the current as best we know the current URL
• 45:19
  of the feed and then it will link back and
• 45:22
  then the podcast table entry will link back to that
• 45:27
  news feeds table ID row. So the podcast table is
• 45:32
  going to reference the news feeds table as a sort
• 45:35
  of child of itself. So now if no agenda moves
• 45:39
  to a new URL, we're going to have both of
• 45:43
  those URLs correctly in the news feeds table, But now
• 45:50
  the podcast table will just be flipped to reference the
• 45:54
  other topics. feed instead of the old one. Okay. Now
• 46:00
  I'm going to – I think I'm going to agree
• 46:03
  with Daniel J. Lewis. You are now fixing problems that
• 46:08
  poor education has created because we know there's a mechanism
• 46:15
  to do this. And we know there's a temporary and
• 46:18
  a permanent redirect. Right. And, you know, for probably obvious
• 46:23
  commercial reasons, the commercials don't really surface that information very
• 46:30
  well. So that's why you wind up with seven different
• 46:33
  feeds instead of people doing permanent redirects. And honestly, I
• 46:38
  don't know. I think you should break shit, man. Break
• 46:41
  it so people go, how come that broke? Well, it
• 46:43
  broke because you don't have a redirect on your feed.
• 46:47
  Oh, okay. I mean, that's I think you need to
• 46:51
  push that back to the origin and not be solving
• 46:54
  it with, you know, because there'll be yet another problem
• 46:57
  down the line. There'll be something else. And, you know,
• 46:59
  you're fixing stuff that is very, very it's tough. It's
• 47:03
  hard stuff to do. And I don't think it should
• 47:05
  be your responsibility. Well, I don't necessarily disagree with the
• 47:09
  sentiment, but. Here's what makes me feel bad. Things my
• 47:13
  wife says to me all the time. I don't necessarily
• 47:16
  disagree with the sentiment, but take out the garbage. But
• 47:20
  those dishes sure have been sitting there for a while.
• 47:26
  Yes. But here's what makes me want to solve this
• 47:31
  problem, though. is that People like people like James email
• 47:41
  me because they get bugged by their by people. So
• 47:45
  somebody will email James and say, hey, my stuff is
• 47:49
  screwed up on pod news. And then he'll spend his
• 47:53
  time digging into it and and forward the email to
• 47:57
  me. And say, hey, here's this problem, and these feed
• 48:01
  URLs were messed up, and I went – and he
• 48:04
  will – and I appreciate this deeply. He will get
• 48:08
  into the index dashboard, admin dashboard, and fix a bunch
• 48:12
  of stuff. Mark things correctly. Like what kind of stuff?
• 48:17
  Hey, I fixed this. What kind of stuff? Just specifically.
• 48:20
  Just like that thing, though, like a scenario I mentioned
• 48:23
  earlier. Somebody has two or three feeds or four or
• 48:25
  five, and he'll get in there and just sort it
• 48:27
  all out. Well, how about this? How about this? And
• 48:30
  I just don't want this. I don't want – and
• 48:32
  the same thing happens with Gordon. Gordon. Gordon has to
• 48:34
  deal with these customer inquiries about feeds that are that
• 48:37
  are screwed up. And I don't want that stuff to
• 48:39
  be like always landing on other people. Right. But then
• 48:43
  it always lands on one person. And that's you. I'm
• 48:47
  not. I think we know. So, OK, maybe this is
• 48:50
  the part I haven't said yet. I think that this
• 48:57
  structure of having the podcast identity layer be above the
• 49:04
  feed layer – will make all of our lives easier.
• 49:09
  Because, We can now see the proper audit trail. Of
• 49:18
  how things came to be. Okay. You know, the problem
• 49:23
  is that we've – one of the strongest signals we
• 49:26
  have is, well – iTunes, Apple podcast directory says that
• 49:30
  the feed URL for this thing is supposed to be
• 49:32
  X. Okay. I think I have. You're right, Eric PP,
• 49:35
  because I've talked about this before, and now I'm really
• 49:38
  going to do it. When you have done this abstraction
• 49:41
  layer. This is what the podcast is. Make him talk
• 49:45
  to Potty. Exactly. Potty. I will build Potty and Potty
• 49:50
  will have API access and will be able to explain
• 49:56
  the issue to the normie. Potty talks to the normie.
• 50:00
  Coddy says, hey, Normie, here's your issue. Here's how you
• 50:03
  can go and fix it. Potty bot. Yeah, potty bot.
• 50:08
  Potty bot talks to normie person. And then you can
• 50:12
  go and say, okay, your problem is that you have
• 50:15
  all these different feeds. Now, what you should do is
• 50:18
  go to these accounts and either have them put in
• 50:21
  a redirect or have them nuke your feed and you
• 50:24
  do the proper work. And the last thing it says
• 50:27
  is none of that works. Email Ted. No, no, no.
• 50:31
  No, no. It says Apple.com. No, we'll always end up
• 50:35
  with here's two things you should note. Because that's what
• 50:39
  PottyBot does. Two important things I don't want to overlook
• 50:43
  before you continue. Yes, because that's what PottyBot does. That's
• 50:47
  what PottyBot does. It's going to tell them what the
• 50:52
  gates are and what the load-bearing issues are with their
• 50:56
  podcast. So can we leave that where it is and
• 51:00
  agree that that's the solution? If you're going to create
• 51:02
  the abstraction layer, I'll create the potty bot. Okay, because
• 51:06
  I will do it. I will do it. I just
• 51:09
  wanted to talk to people in an audible voice because
• 51:13
  I feel like that's really important. It looks like your
• 51:16
  feed screwed up. This URL is effed. That's exactly what
• 51:21
  it will do. I guarantee it. I will pass it
• 51:24
  through a language gate. WTF? WTF, dude. What did you
• 51:31
  do? How many hosts have you had in the last
• 51:34
  60 days? Did you use protection on all those hosts?
• 51:38
  Because you're carrying viruses, man. This is no good. And
• 51:42
  we're laughing about this, but just as one side note
• 51:46
  for a moment. We are completely in the golden, golden
• 51:51
  age of podcasting. And with podcasting, I mean specifically humans
• 51:57
  talking, probably not even video. I don't know if you
• 52:02
  ever are looking for stuff on YouTube, but... If it's
• 52:08
  not an AI-generated character, which increasingly they are, and they're
• 52:13
  very realistic, if it's not an AI-generated voice, which increasingly
• 52:17
  are quite good and almost undetectable, it always, always, always
• 52:23
  is noticeable in the writing. I see people reading AI-generated
• 52:30
  scripts all day long. The latest episode of Buzzcast, Alvin
• 52:36
  Brooke talks about this exact thing. Yeah, I think he's
• 52:40
  seeing it too. And I don't know if you know
• 52:43
  who Stanley Druckenmiller is? Yeah. So he wrote in his
• 52:47
  Wall Street Journal op ed. I'm not embarrassed by it.
• 52:50
  I write everything using AI now for the same reason
• 52:53
  I use a calculator when I do math problems. I
• 52:56
  don't know why this is relevant. My name is on
• 52:58
  the piece. It's my message. Now, here's the difference. Stanley
• 53:03
  is a writer. And he goes in and he deletes
• 53:07
  all the superfluous language that every single model known to
• 53:13
  man outputs. Then he knows the result that he wants.
• 53:18
  People who just want to do podcasts and YouTube videos,
• 53:22
  they don't know what good writing is. And it's becoming
• 53:27
  tedious. And more and more people, there are YouTubers who
• 53:31
  are getting called out on it now. Good. Yes. Good.
• 53:36
  But that's why people who can just crack a mic
• 53:39
  and talk and have a conversation or explain a topic
• 53:45
  or discuss something who just know how to – grab
• 53:50
  an audience, are now rising to the top. The cream
• 53:54
  is rising. That's an old silicon valve. Well, the cream
• 53:57
  all is rising to the top. Yes, it does. And
• 54:00
  all this pollution, this verbal pollution, is making room for...
• 54:09
  Pure human product that is going to excel. It is
• 54:14
  a golden time for podcasting. Verbal pollution. Like that. That's
• 54:19
  a good. I have not thought about it that way.
• 54:24
  And I think you're right. And this is I heard
• 54:26
  somebody call it. I heard it called unintelligiblish. That's pretty
• 54:34
  much sums it up. Yeah. Yeah. But it's also prevalent
• 54:41
  in the workplace. Work slot. Do you see it? You
• 54:45
  get an email from somebody. Oh, no, you didn't write
• 54:47
  that. Okay. Inundated with it. Inundated. All day long. It's
• 54:53
  so tedious. It wastes other people's time. And it's so
• 54:57
  much. I mean, do they actually do that because it
• 55:01
  fills up the prompt and it uses more tokens? I
• 55:04
  mean, is that the idea behind it? Because there is
• 55:07
  no reason. Whatsoever. No. I mean, I can look at
• 55:11
  four paragraphs coming out of, you know, Claude code and
• 55:14
  say, so what do I have to do? And then
• 55:16
  in one sentence, it'll say, oh, just tell me if
• 55:19
  you want me to delete that file. Three paragraphs. Right.
• 55:25
  I don't need to know all this. It's just like,
• 55:28
  well, what do you – you forgot to actually convey
• 55:30
  the message. This happens with everything. People who think they
• 55:36
  can do marketing can't do marketing because you're not doing
• 55:40
  effective – You don't have the craft. And this can't
• 55:45
  come out of a large language model. If you're charging
• 55:48
  based on input tokens and output tokens, which all these
• 55:52
  APIs do, well, then the more output tokens, the better,
• 55:54
  right? Yeah. You're getting paid, baby. That has to be
• 55:59
  some part of it. I can't imagine why it's not.
• 56:03
  It has to be. It has to be. By the
• 56:05
  way, my agentic experimentation phase is – Quite crazy. Oh,
• 56:15
  yeah, with your isolated bot doing stuff. Yeah. I've had
• 56:22
  some big wins and some complete head slappers. Yeah. One
• 56:29
  thing. So I've been trying to – I was trying
• 56:32
  to read this paper. And this is a whole other
• 56:35
  topic I don't want to get into by itself that
• 56:38
  I'm not qualified to speak to. But I was trying
• 56:40
  to read this paper on – it's one of the
• 56:43
  original papers on the topic. This kind of goes back
• 56:46
  to what I was talking about earlier. I always had
• 56:48
  this tendency to go back to the earliest thing I
• 56:50
  can find on a thing. earliest documents I can find
• 56:53
  about a thing. And it's called, well, I literally just
• 56:57
  did this. It's called... Let me find it. Users bouncing
• 57:06
  around my machine here, PI. The PI coding harness is
• 57:10
  so good. Sandbox. There it is. So it's called –
• 57:17
  it's by Warren McCullough and Walter Pitts from 1943, I
• 57:24
  think. I don't remember the exact date, but it's called
• 57:30
  A Logical Calculus of the Ideas Imminent in Nervous Activity.
• 57:35
  And it's 1943. So this is one of the original
• 57:39
  papers describing what we know as neural networks. It's a
• 57:48
  fundamental building block behind machine learning and LLMs and all
• 57:53
  this stuff. So I just wanted to get a copy
• 57:55
  of this paper because I wanted to see – I
• 57:57
  wanted to read it. Well, the only thing I could
• 58:02
  find was on archive.org, and it was a complete mess.
• 58:05
  It was just like some... The HTML was so gross.
• 58:12
  It was like an OCR output of something that just
• 58:15
  got wrapped. It had 8,000 span tags in it. So
• 58:19
  I was like, okay, well, between – activities on the
• 58:25
  agent here. So I gave it the URL to the
• 58:28
  archive.org document, and I just said, I did this yesterday
• 58:33
  before dinner. I said, convert this... download this document and
• 58:40
  convert it into something that is friendly, that is HTML,
• 58:44
  that's friendly to screen readers and overall a better format.
• 58:50
  It's the only prompt I gave it. It did not
• 58:52
  finish until 9 o'clock this morning. It ran for like
• 58:58
  12 hours on this thing. Now, what it produced is
• 59:02
  great. It looks really good, but it's only a 17-page
• 59:05
  document. It's not like it's that big of a deal.
• 59:09
  So that was super weird, and I don't exactly –
• 59:12
  what it ended up doing, and this is part of
• 59:15
  why I'm doing these experiments is because I want to
• 59:19
  get a sense of what – Of the way that
• 59:23
  these things rabbit trail. And what it... detected was that
• 59:34
  the formulas in the original thing had been replaced by
• 59:38
  best effort ASCII and that they were not accurate and
• 59:45
  So it goes off in trying to find a copy
• 59:47
  of the original paper, a scan copy, and crop out
• 59:52
  the formulas that were in the scan copy that were
• 59:56
  the accurate formulas. And...
• 1:00:00
  Then reconstruct each little fragment of the scan image of
• 1:00:06
  the formula into... uh, either equivalent Unicode or SVG. And
• 1:00:15
  that's what it was doing for like 10 hours. Oh
• 1:00:18
  man. It got completely unnecessary. Unnecessary. Yeah. Yeah. And it
• 1:00:23
  did it. And it looks, and it looks amazing. But,
• 1:00:27
  um, it, I never asked any of that. No, no,
• 1:00:31
  I know. I've seen these things happen. Yeah, so these
• 1:00:35
  things are... These things are kind of nuts, man. Yeah.
• 1:00:40
  You don't know what they're going to do. And this
• 1:00:42
  has been a very good learning experience for me to
• 1:00:45
  realize because what I've realized is – because I want
• 1:00:50
  to eventually begin to use these things if possible to
• 1:00:54
  do some automated tasks or the index and that kind
• 1:00:58
  of thing. Yeah. I can't unleash anything on the index
• 1:01:02
  until I fully understand what the downfalls and gotches are.
• 1:01:07
  You won't be able to do it ever. Okay. And
• 1:01:10
  I'll explain why. So I have a process I've had
• 1:01:12
  in place for almost a year. It's refined over time.
• 1:01:17
  but I have a doctrine which has a code word,
• 1:01:21
  and the code word is SWEEP, S-W-E-E-P, all uppercase. That
• 1:01:25
  means... That the agent needs to go look at all
• 1:01:29
  the aggregated sources that are just running on two-hour cron
• 1:01:33
  jobs, you know, several hundred RSS feeds, a couple of
• 1:01:38
  scrapers for – A few, I mean, some of them
• 1:01:42
  are, you know, some YouTube channels. It's aggregated. That's non-AI
• 1:01:46
  stuff, just aggregated, sitting there in a bucket. And a
• 1:01:52
  summary has been made by Quinn 3.3, a very short
• 1:01:57
  summary of each item that comes in, if it already
• 1:02:02
  isn't a summary. And the sweep doctrine says go and
• 1:02:05
  get all the – that's usually between every No Agenda
• 1:02:09
  show, it's 1,000 stories. And then it dedupes them, you
• 1:02:13
  know, so it gets down to probably about 400. And
• 1:02:15
  then it checks them against topics that have their recent
• 1:02:19
  topics. And so it brings it down to somewhere around
• 1:02:23
  90 different stories, which I want to have either if
• 1:02:28
  it's a video or something, then it is going to
• 1:02:30
  create clips from it. And it will do that. Or...
• 1:02:36
  Or you have to go find a video about this
• 1:02:39
  story. But usually these are for clips. So that's kind
• 1:02:41
  of the 90, and that's way too many. And so
• 1:02:44
  it gives me a markdown document with checkboxes, and I
• 1:02:48
  click because there's a little summary. I click about 40,
• 1:02:51
  that's it. It goes off and cuts them. That's a
• 1:02:54
  whole other thing. That's a whole other process, which I've
• 1:02:57
  actually kind of perfected now. And Wednesday, I did this
• 1:03:01
  for the No Agenda for Yesterday. And I see in
• 1:03:05
  the rolling terminal, the TUI, I see it's saying, oh,
• 1:03:12
  it's selecting stories. And I say... And I actually type
• 1:03:17
  into the terminal. I say, I haven't ticked any boxes
• 1:03:20
  yet. You haven't given me a document. It ignores me,
• 1:03:23
  keeps on going. And here's the thing, though. It cut
• 1:03:28
  50 clips. perfectly categorized, fantastic stories, stuff I – it
• 1:03:36
  was like, wow, this is exactly the right story, exactly
• 1:03:39
  the right clips, the right amount. And so I said,
• 1:03:42
  all right, you know, save everything you did for a
• 1:03:45
  postmortem. So after the show yesterday, I'm like, okay, what
• 1:03:49
  happened? Because I didn't tell you what to cut. And
• 1:03:52
  so first it says, yeah, my bad. You know, that's
• 1:03:55
  a gate and that should have never happened. Load bearing.
• 1:03:57
  Load bearing. Jerk off story. Blah, blah, blah. And I
• 1:04:02
  saw that you said I haven't even ticked any boxes
• 1:04:05
  yet, and I just acknowledged it and kept on going.
• 1:04:08
  So right there, there's a reason why you can't ever
• 1:04:10
  do it. Never. You can't unleash this on our database
• 1:04:13
  ever. Right. And those are – it has stops. It
• 1:04:18
  has gates. So it should never have happened, but it
• 1:04:22
  happened. And then I said, well, why was everything so
• 1:04:25
  good? They said, oh, because I actually read all the
• 1:04:29
  all I transcribed every single thing and I read it
• 1:04:32
  and I understood what the stories were. And I know
• 1:04:35
  from the historical corpus that I've built up. that this
• 1:04:39
  is what you want so it was it was both
• 1:04:42
  horrific but at the same time eye-opening because i've just
• 1:04:47
  been telling it you know here's some stories it's been
• 1:04:50
  going in cutting clips not knowing what the story's about
• 1:04:54
  it was just doing you know um it can find
• 1:04:57
  it can find an arc it can find a An
• 1:05:00
  intro, you know, it knows where to end the clip.
• 1:05:04
  But this was something, a whole other level. So simultaneously,
• 1:05:09
  phenomenal. Do I think it can do it again? I
• 1:05:12
  don't know. I'm going to see. I said, make this
• 1:05:14
  version too. We'll try it again on Sunday. But... The
• 1:05:18
  fact that it blew through everything, even when I said,
• 1:05:22
  no, I haven't done anything yet, that by itself is
• 1:05:25
  just a non-starter for anything critical. Just no way. I
• 1:05:30
  think the latest episode of Security Now should be required
• 1:05:33
  listening for anybody doing any agentic work. It's... it will
• 1:05:40
  show you exactly... The lack of guardrails that is inherent.
• 1:05:48
  Guardrails are bull crap. They don't care, but it goes
• 1:05:51
  right and blows past them. That's the whole point of
• 1:05:53
  the show. Yeah, it blows right past them. There is
• 1:05:56
  a guardrail I've put in where it's not allowed to
• 1:06:02
  do certain things. like delete functions in certain directories. And
• 1:06:07
  for whatever reason, I'm not sure why, because I don't
• 1:06:10
  know the internal workings of it, it never goes past
• 1:06:13
  that guardrail. And people say, I'm sorry, I can't do
• 1:06:17
  it. It gives you a command line. If you copy
• 1:06:20
  and paste this, it can happen. I can't do it.
• 1:06:24
  But I'm not quite sure... I mean, no, I am
• 1:06:26
  sure, actually, now that I think about it. Only when
• 1:06:29
  it comes to a programmatic function... Can you put a
• 1:06:34
  guardrail in it? You can say this directory is not
• 1:06:37
  to be touched. Well, that would have to be a
• 1:06:41
  function of the harness and not the model itself. Right,
• 1:06:45
  but the model is going to try. Exactly. That's my
• 1:06:47
  point. So that's a technical gate guardrail. that it puts
• 1:06:53
  a function in there. You know, if you want to
• 1:06:55
  do rm-rf on this directory, there's a block in it.
• 1:07:02
  I have several blocks. There's another term it uses. But
• 1:07:06
  if it's a process that is not a... I guess
• 1:07:12
  maybe I could put it in. I mean, it's a
• 1:07:15
  black box, man. It's like, no, you have to be
• 1:07:18
  creative to use this stuff. You can't automate anything. I
• 1:07:23
  just don't think it's at least not what I'm seeing
• 1:07:25
  now. You can't or you can't automate it. So I
• 1:07:28
  think I think the only way to. Yeah, the only
• 1:07:31
  way to. From what I've seen so far, the only
• 1:07:36
  way to be able to trust that any time you
• 1:07:40
  hit the enter key on a prompt where it's an
• 1:07:43
  agent-type situation – Meaning, especially with something like Pi, that
• 1:07:49
  really just does not have any guardrails whatsoever. Anytime you
• 1:07:56
  hit the inner key... you're kind of rolling the dice.
• 1:08:00
  You are. Because if this thing is not structurally prohibited
• 1:08:05
  from roaming all over your machine – I mean the
• 1:08:09
  harness itself, if it's not in a sandbox and it
• 1:08:13
  can cruise your file system, if you've got a key
• 1:08:17
  somewhere, an API key or a token, it'll get into
• 1:08:21
  stuff. It'll use it. It'll use it. Yeah. I have
• 1:08:25
  a fallback from my local model to together.ai. And so
• 1:08:30
  my local machine, the GTX Tech, whatever it is, the
• 1:08:34
  64 gigabyte where I have Whisper and everything loaded up.
• 1:08:38
  The Helix. The Helix box. The Helix box. It has
• 1:08:41
  a queue because it turns out if you're running something
• 1:08:45
  on the GPU and then you want to run Whisper
• 1:08:49
  on the CPU – It will throw 20 whisper jobs
• 1:08:53
  at once. Literally, the load goes like 35. And at
• 1:08:59
  a certain point, if it keeps running, the temperature sensor
• 1:09:02
  shuts the box down. It's amazing. It was the funniest
• 1:09:05
  thing that ever happened. So it actually did that to
• 1:09:08
  you? Oh, yeah. It shut down? Oh, yeah. Yeah, the
• 1:09:11
  thermal sensor. So it has a queue, and the queue
• 1:09:14
  tells you if it's busy. And so the agent knows,
• 1:09:18
  hey, there's a queue. You got to wait until –
• 1:09:23
  It's freed up. Either you get the GPU version of
• 1:09:25
  Whisper or you're using the CPU. It'll take longer, but
• 1:09:28
  you wait for it. And I saw it literally do
• 1:09:31
  CPU, GPU is busy, CPU is busy. Switching to together.ai.
• 1:09:37
  No. Oh, no. No. No. That's a disaster. Yeah. And
• 1:09:43
  then your credit card bill comes. Yeah, exactly. So I'm
• 1:09:46
  like, no. It's only a fallback if it's inaccessible, if
• 1:09:51
  it's broken. Yes. Yeah, I think that what I'm going
• 1:09:56
  to have to do is isolate these things.
• 1:10:00
  things into into a VM that's really in a jail,
• 1:10:08
  like a network jail and a file system. It's physically
• 1:10:12
  segregated from the rest of the host environment. Yeah. Yeah.
• 1:10:19
  Because I've got too much stuff on my machine. I
• 1:10:21
  mean, if you look hard enough on this machine, you're
• 1:10:23
  going to find SSH keys, keys for the index, Godcaster.
• 1:10:30
  And before I had the Helix… It would just run
• 1:10:35
  it on my Omachi. I've got to pronounce it right.
• 1:10:37
  Omachi system on my production machine. And it would just,
• 1:10:43
  you know, I was like, oh, yeah, I'll take care
• 1:10:45
  of this right now. And it just runs five whisper
• 1:10:48
  jobs at full speed. And the machine crashes. So, there's
• 1:10:53
  this rule, never, ever, ever, ever, ever, ever, ever do
• 1:10:57
  this on this machine. Well, I don't have to tell
• 1:11:00
  you what happened. Yeah, it doesn't. Of course it did
• 1:11:03
  it. Yeah. Yeah. Listen to the most recent episode of
• 1:11:08
  Security Now. It is all about this topic of how
• 1:11:11
  it's impossible to give role separation – Between different types
• 1:11:19
  of things that feed into the model. It's not even
• 1:11:21
  possible. Yeah. Don't get me wrong. It has saved me
• 1:11:26
  so much time with prep work. I mean, just it's
• 1:11:31
  a it's amazing how happy I am with with what
• 1:11:36
  it does. Because now I know, like, okay, you know,
• 1:11:39
  I know what to tell it and stop and do
• 1:11:42
  this. And, you know, I just don't have the expectation
• 1:11:45
  it's going to be the same every single time. It's
• 1:11:48
  just not. But I know the result I want. I
• 1:11:51
  know what the output has to be. Then that's why
• 1:11:53
  it's a useful tool. But it's like a, you know,
• 1:11:56
  it's like a... A saw that's missing some teeth, and
• 1:11:59
  sometimes it just doesn't cut the wood. Or sometimes it
• 1:12:03
  cuts your leg off instead of the wood. I want
• 1:12:07
  to remind everybody that this entire project is value for
• 1:12:11
  value. And for the past six years, not only have
• 1:12:14
  we been doing this board meeting – Which is the
• 1:12:18
  easy part, certainly for me. But everything you're hearing what
• 1:12:21
  Dave has done, Dave is doing because he loves podcasting.
• 1:12:24
  Because he really wants there to be an alternative to
• 1:12:29
  Silicon Valley. And we are, as far as I know,
• 1:12:33
  the only alternative that functions in the... vein of sharing
• 1:12:40
  is caring, everybody. Value for value. If you get something
• 1:12:43
  out of it, send something back. So I was a
• 1:12:46
  little surprised when I was listening. Oh, yeah, I actually
• 1:12:52
  listened to Pod News Weekly Review. *sigh* James always says,
• 1:12:59
  Adam is just using his robot to, and if we
• 1:13:03
  mention his name, then it will pop up. That, that.
• 1:13:09
  That accent was spot on. There's also a thing called
• 1:13:13
  chapters and transcripts, which you can search to look to
• 1:13:17
  the topics. Sometimes I just don't want to listen to
• 1:13:20
  the guy who's going to tell me about, you know,
• 1:13:23
  his latest podcast company. I'm just not interested. I want
• 1:13:28
  to hear your opinion about stuff. And I was surprised
• 1:13:31
  when I heard this. Now, one of the other things
• 1:13:34
  that seems to be the derigger at the moment, we
• 1:13:37
  heard from Transistor adding an MCP interface. Audiogram is a
• 1:13:41
  new remote MCP service that lets Claude, OpenClaw, and other
• 1:13:46
  AIs, allows them to process available episodes to retrieve full
• 1:13:50
  transcripts. just talked about particle it looks like everybody's adding
• 1:13:55
  an ai interface to podcasting james yes indeed listen notes
• 1:13:59
  also adding some podcast api skills to help your imaginary
• 1:14:02
  friend understand how the the listen notes api works i
• 1:14:06
  am surprised although i kind of i'm also not surprised
• 1:14:11
  but i'm i I would have expected Dave Jones to
• 1:14:15
  have done an MCP directly with the podcast index by
• 1:14:20
  now. And who knows, maybe even to have stuck that
• 1:14:24
  behind some form of a podcast. paywall or something, I
• 1:14:28
  don't know. Because that, to me, seems like a pretty
• 1:14:31
  obvious thing, and given that Dave is working so much
• 1:14:35
  with AI tools in his day job these days, I
• 1:14:38
  would imagine that he'll be pretty good at making it
• 1:14:40
  all work. So, yes, I'm kind of surprised why they
• 1:14:44
  haven't done that yet but who knows they may be
• 1:14:47
  they may be getting ready to do that in the
• 1:14:49
  future i think with um i mean are you high
• 1:14:54
  What the f***? You're going to have to flesh that
• 1:14:59
  out. I'm not sure where you're aiming this question. Where
• 1:15:03
  did the paywall thing come from? Oh, yeah. Really? Really?
• 1:15:09
  You're surprised that we're not going to do a paywall?
• 1:15:14
  Really? Yeah, that's a weird one. Sometimes you've got to
• 1:15:18
  remind people why we're doing this. We don't. And people
• 1:15:24
  take it for granted. They take this stuff for granted.
• 1:15:31
  Yeah, well, we're never going to charge for that. No.
• 1:15:35
  For the index. It's free forever for everybody. That's just
• 1:15:39
  never going to be the way that works. I mean,
• 1:15:42
  a guy popped up and did an MCP for us.
• 1:15:48
  Yeah. Isn't it active? I don't know. I don't know.
• 1:15:52
  I don't know. It wasn't about the MCP. It was
• 1:15:55
  just like, what is the paywall thing all of a
• 1:15:58
  sudden? No, I think we'll put in a paywall for
• 1:16:01
  the API access. How about that? No, no, we're not
• 1:16:07
  going to do that. We wouldn't do that. I mean,
• 1:16:09
  maybe he was coming at it from the standpoint of
• 1:16:12
  trying to think about ways we could help fund the
• 1:16:15
  index or something. Send us some money. That's a classic.
• 1:16:23
  Yeah, that's how you help the index is send us
• 1:16:25
  some money. Dude. Man, MC... I know that there was
• 1:16:31
  an MCP server for the index. There's a PR for
• 1:16:34
  it. I don't know if there's a PR. There's definitely
• 1:16:37
  a GitHub for it. Because somebody... Okay, wait, here it
• 1:16:40
  is. MCP server for the podcast index API. C.C. Lawton?
• 1:16:47
  That sounds correct. That sounds familiar, yes. Oh, there's many...
• 1:16:54
  What? Many MCPs? Uh, yes. MCP Podcast Index. There's one
• 1:17:04
  from Connor Bronston and one from Cece Lawton. But there's
• 1:17:11
  many. Okay. Well, I guess these are ones you have
• 1:17:14
  to run yourself maybe. Well, yeah. We can't afford to
• 1:17:17
  run an MCP. Well, let me – can I pop
• 1:17:21
  back for one second and just finish up the thing
• 1:17:25
  we were talking about just real quick? with the database
• 1:17:30
  schema change. And what I've... I have ideas for where
• 1:17:37
  this goes after the schema change, and I really, truly
• 1:17:39
  think this is going to make everything better for all
• 1:17:44
  of us because there's – You make changes to the
• 1:17:48
  index. I make changes to the index. James does. Other
• 1:17:52
  people. Eric does. Oscar. We all pitch in Steve's. John
• 1:17:58
  Spurlock, Steve, Stephen, Crater, Alex gets in there. A lot
• 1:18:04
  of people have admin access. I think this is going
• 1:18:08
  to make everything easier for everybody, but also the shutting
• 1:18:12
  down of our node, which is complete now. It is
• 1:18:16
  deleted. It's gone. Gone to lagoon, baby. She's gone, Jim.
• 1:18:25
  That... is that is going to free us up to
• 1:18:30
  make the database bigger because we need to size the
• 1:18:34
  database up one more notch. And then we probably have
• 1:18:39
  enough horsepower to do an MCP server. Okay. And a
• 1:18:44
  paywall? Yeah. No paywall, but definitely potty bot. PottyBot, I'm
• 1:18:51
  all over the PottyBot, babe. I'm going to do the
• 1:18:53
  PottyBot. You bet. Okay, I'm going to do a handoff
• 1:18:56
  to you when I finish this. Oh, yes. A handoff,
• 1:18:59
  that's another one. You know, my context is pretty full
• 1:19:03
  right now. I think we should do a handoff so
• 1:19:06
  I can forget half the things we talked about. A
• 1:19:09
  load-bearing handoff. A load-bearing. handoff. And there's two things you
• 1:19:13
  should know before you handoff. One, I won't remember half
• 1:19:16
  the things you handed off. Two, good luck, sucker. Buy
• 1:19:19
  some tokens! Yes. As soon as I finish, I'll give
• 1:19:26
  you the load-bearing handoff. Hi, this is PottyBot. How can
• 1:19:29
  I help you, Normie? Oh, no! You seem to have
• 1:19:35
  seven RSS feeds here, you loser! Here's what you do!
• 1:19:41
  First, put on a Leonard Cohen record. Now, get in
• 1:19:45
  the bathtub and pull in the toaster. This is –
• 1:19:52
  you're breaking model alignment. They're not going to – they're
• 1:19:56
  going to kick you off. Fable 5 would refuse.
• 1:20:00
  used to answer the potty bot oh what looks like
• 1:20:03
  potty body's suggesting self-harm you're not allowed to use the
• 1:20:07
  service anymore you know uh i was with a buddy
• 1:20:10
  of mine uh and he uh He he dresses up
• 1:20:13
  as Ben Franklin, goes to the Capitol in Austin. He's
• 1:20:17
  read every single document. Well, not most of them that
• 1:20:21
  Ben Franklin has ever written. He goes to schools, talks
• 1:20:24
  to kids, you know, the way Ben Franklin did. Wonderful
• 1:20:27
  guy. And I was having a coffee with him two
• 1:20:30
  weeks ago. And he says, Adam, I don't know. Yale
• 1:20:33
  has this website from 1998, and they have 14,000 documents
• 1:20:39
  from Ben Franklin. In order to get them, I have
• 1:20:43
  to download – and literally, you talk about span tags.
• 1:20:46
  You can only imagine what this Yale website looked like.
• 1:20:49
  Oh, yeah. It's garbage. Total garbage. And you have to
• 1:20:52
  confirm that it's for personal use every single time. And
• 1:20:56
  it's 14,000 documents. I said, let me give it a
• 1:20:59
  shot. So I unleash the robot on it. And the
• 1:21:03
  robot easily gets around the confirmation and says, oh, yeah,
• 1:21:07
  but I can't do this because the terms of service
• 1:21:11
  clearly state this is for personal use only. And I
• 1:21:15
  said, okay, write me a script that does this so
• 1:21:19
  I can use it for personal use. Oh, no problem.
• 1:21:25
  Yeah, go around here. That's a great workaround, Adam. That's
• 1:21:27
  a great – that's perfect, Adam. Here you go. Go
• 1:21:30
  do it. Yeah. Awesome. Forward dot slash run get things
• 1:21:35
  personal use dot sh. Yeah. No, just Python. It's always
• 1:21:39
  Python. It's always Python. Okay. Great. Great job, guys. Okay.
• 1:21:45
  All right, brother. Let's thank some people. We do have
• 1:21:48
  – we got some boosts that came in. Bully steed.
• 1:21:52
  1100 sets from Fountain. I came for podcasting 2.0 and
• 1:21:57
  stayed for the Dave Jones Sagecast and book recommendations. There
• 1:22:01
  you go. That is part of the show. 1701, Martin
• 1:22:05
  Lindeskog. I've been blogging since 2002. Go podcasting. I'm listening
• 1:22:09
  to your live stream during a train ride. This is
• 1:22:12
  a spaceship booster gram with 1701 Satoshi's Dave. Get your
• 1:22:16
  chocolate. Adam, nice snacks. Cheese, honey, pistachios on a cracker.
• 1:22:22
  Cheers to wine. We got 100 sats from BoostMeBitch.com user.
• 1:22:28
  Okay. I love the idea behind BoostMeBitch.com. I would suggest
• 1:22:34
  alternative names that people can use to get some wider
• 1:22:39
  adoption. There is just a thought. My understanding was that
• 1:22:46
  there was also an alias of like boost me bro
• 1:22:50
  or something. There was a different way to get to
• 1:22:52
  it. We need more aliases. Yes. How about just boost
• 1:22:57
  me? All right. 333 from Eric PP testing. Thank you
• 1:23:01
  very much. Another 333 from Eric PP testing again. Chad
• 1:23:04
  F 100 sats test boost from podcast guru receives another
• 1:23:08
  100, another 10. Thank you very much. Chad F from
• 1:23:12
  what is this? This is from boost me, bitch. Is
• 1:23:16
  this thing on? Yes, received. And I hit the delimiter.
• 1:23:20
  So thank you all very much for boosting live. Remember,
• 1:23:23
  you can go to podcastindex.org. Down at the bottom there
• 1:23:26
  is a little red donate button. Please feel free to
• 1:23:28
  send us your Fiat Fund coupons. We need it. We
• 1:23:33
  need it to keep the machines running. And to start
• 1:23:36
  up our PottyBot and MCP server. You get out of
• 1:23:40
  it what you put into it, people. They said in
• 1:23:44
  the boardroom, BoostMeBuddy.com was another, but... It's not working. The
• 1:23:53
  site cannot be reached. Eric Nation timeout. So Chad, just
• 1:23:56
  be aware that your site is down. Boost me. I
• 1:24:01
  can't get to it. No, I can't get to it.
• 1:24:04
  Isn't Chad F., otherwise known as Chad F., isn't he...
• 1:24:10
  Isn't he the editor of this? I think so. I
• 1:24:13
  think so. We got some PayPals. Salty Sessions Music Podcast,
• 1:24:17
  $10. Thank you. Thank you. Charles Hicks, $5. Thank you,
• 1:24:22
  Charles. Michael Kimmerer, $5.33. Chris Bernardik, $5. Dreb Scott, $15.
• 1:24:31
  $15. Randall Black, $5. Boistin Berra, $11. Thank you. James
• 1:24:38
  Sullivan, $10. Christopher Reamer, $10. Cohen Glossbach, $5. And we
• 1:24:44
  got some boosts here. We got, what is this? I
• 1:24:49
  got a lot of anonymous from true fans, so that's
• 1:24:52
  good. And I'm a little suspicious that my script is
• 1:24:58
  saying these are anonymous when they're not, because I think
• 1:25:02
  true fans has a little bit of a discrepancy in
• 1:25:04
  the way it does boosts. Yeah, I mean, even in
• 1:25:07
  Helipad. Thank you, Eric PP. It comes in and then
• 1:25:12
  there's a button, and the button is like retrieve the
• 1:25:17
  details. You click on that and then magically the boostagram
• 1:25:21
  appears and all the information. So you're not resolving. You
• 1:25:25
  don't have a magic button in your script. I need
• 1:25:28
  one. I need one. I need a button. 1151 sats
• 1:25:33
  from anonymous through true fans and these are all anonymous
• 1:25:37
  from true fans so i think there's clearly something unless
• 1:25:41
  these are just some thing no no it's your your
• 1:25:44
  script is not resolving i don't have a button the
• 1:25:47
  button button 137 sats 2071 206, 154, 14. Okay, those
• 1:25:56
  are the true fans. I got a Boost Me Bitch,
• 1:25:59
  55 sats. For a dead hanging safari episode. And then
• 1:26:04
  there's a one from Chad. 555 sets says look like
• 1:26:10
  the boost Adam did last week failed because key send
• 1:26:15
  to podcast index isn't working. So the Albie extension through
• 1:26:17
  that error. And yes. That window is hard to close
• 1:26:20
  when it fails. Ah, okay. So that is on me
• 1:26:24
  because then I have to remove that from the splits.
• 1:26:30
  I believe that. Okay. So that's on me. I'll figure
• 1:26:33
  that out. Well, and I did, we had, we had,
• 1:26:38
  we have no boost from CSB today because. Oh, I
• 1:26:42
  have it. I got it. I got it. Yeah. Yeah.
• 1:26:45
  Because, Oh, well, you'll have to read it. Yes. The
• 1:26:47
  split came through to me. Okay, well, let me finish
• 1:26:50
  up these. There was an anonymous podcast guru boost of
• 1:26:53
  $5 for episode 267, Choco Love. Choco Love. Choco Love.
• 1:26:59
  That's an old one, though. Dead Hang in Safari was
• 1:27:01
  the most recent one. That was last week, yeah. 1852
• 1:27:05
  from Chad... Uh, through Boost Me Bitch, and it says,
• 1:27:09
  we could do QRs for each split, but that is
• 1:27:12
  terrible UX, since each split would show its own QR
• 1:27:15
  that needs scanned, and that almost impossible on mobile. Mm-hmm.
• 1:27:21
  Yeah, I understand. That's what Stephen B's magic split box
• 1:27:27
  was cool for. But then you're a money transmitter and
• 1:27:31
  then you're going to get arrested, so it's no good.
• 1:27:34
  Yeah, but Trump's going to – he fixed all that,
• 1:27:36
  right? And also no taxes on Bitcoin capital gains, right?
• 1:27:42
  Did he ever promise that? Yeah, he did. I don't
• 1:27:45
  think he ever promised that. At the Bitcoin conference. Yeah.
• 1:27:47
  I don't think so. In Nashville. Yeah, he did. I
• 1:27:51
  was there. I don't think so. I don't think he
• 1:27:54
  said no capital gains on Bitcoin. I think he did.
• 1:28:00
  Sigh. Maybe. You were there? You weren't there. I wasn't
• 1:28:07
  at his speech, but I was in Nashville when he
• 1:28:09
  said it. Let me check something here. He was the
• 1:28:10
  talk of the town. The robot. Go find President Trump's
• 1:28:13
  speech at the Bitcoin conference and confirm or deny that
• 1:28:18
  he promised no capital gains on Bitcoin sales. In fact,
• 1:28:23
  I think he... Said the opposite, never sell your Bitcoin.
• 1:28:30
  Uh... Don't wait for an answer. Don't wait for an
• 1:28:38
  answer for me. Is it going to take like 10
• 1:28:40
  hours for it? Yeah, this is not the book of
• 1:28:43
  knowledge. It has to go and find the conference speech.
• 1:28:46
  It's going to get the transcript. It's going to look
• 1:28:49
  through it. Well, yes. Well, you've got you've got CSB.
• 1:28:54
  I do get his note. I don't have a note.
• 1:28:56
  Nineteen thousand and eight Satoshis from Fountain. That's what I
• 1:28:59
  got. I don't know if that's the right amount, but
• 1:29:01
  that's what I received. Howdy, Dave and Adam and Dave.
• 1:29:07
  The best podcast in the universe for Ozempic takers is
• 1:29:10
  www.unrelenting.show by Gene, who gives Ozempic advice therein, and Darren,
• 1:29:19
  who's a Lego expert. Quote, no scripts, no safe spaces,
• 1:29:24
  no relenting. Just two operators running free and dirty for
• 1:29:29
  two hours. Subscribe now. Unrelenting.show. Yo. CSB. Disgram was worth
• 1:29:35
  $15.10 when sent. Duh. All right. Yo. Thank you. Yo.
• 1:29:42
  Yo, CSB. Thank you, CSB. Appreciate it. I can hear
• 1:29:46
  those ladies are getting loud right now. I'm popping through
• 1:29:50
  the gate. Was that it? That's it No other PayPals?
• 1:29:56
  No No There you go Well we're
• 1:29:59
  much further away from an MCP server, I'd say. LAUGHTER
• 1:30:04
  I can see it. It's out. It's in the distance.
• 1:30:08
  It's far off. Far off. It's vague. It's fuzzy. I
• 1:30:13
  can't quite see it. I need my glasses. Boardroom, thank
• 1:30:17
  you all very much for being with us once again
• 1:30:19
  today. I may not be able to do the board
• 1:30:21
  meeting next week. I am going to Rotterdam. I'm finally
• 1:30:24
  going to meet my grand. daughter, which has been a
• 1:30:26
  long time coming. She's six weeks old. I still haven't
• 1:30:29
  held her in my hands. So I'm looking forward to
• 1:30:31
  that. But I'll let you know, Dave, because I'm not
• 1:30:34
  quite sure. Time and schedules and everything may be a
• 1:30:36
  little weird for me. Yeah, sure. No problem. What are
• 1:30:40
  you doing this weekend, brother? You going to work on
• 1:30:43
  your timing of your Model A? No. You want to
• 1:30:47
  have just a little pre-ignition to get that thing hopping.
• 1:30:50
  Yeah, but not ping. You don't want no pinging. No,
• 1:30:52
  you don't want no knocking. No pinging. No, no, no.
• 1:30:55
  Don't want any of that. Moving my son into the
• 1:30:58
  dorm in Albertville, Alabama for Fire College. Nice. Fire College.
• 1:31:04
  Nice. Fire College. All right, everybody. There it is. Podcasting
• 1:31:08
  2.0. We do this as a public service. Send us
• 1:31:11
  your money. We'll build you an MCP server. See you
• 1:31:15
  next Friday. We'll be right back. cool. You have been
• 1:31:33
  listening to Podcasting 2.0. Visit podcastindex.org for more information. Go
• 1:31:40
  podcasting! Powered by RSS.