WEBVTT 00:00:00.000 --> 00:00:04.620 Hello, in the Crown of Aragon 639 years ago, 00:00:05.059 --> 00:00:11.080 it was March 29th, 1387. My name is Jonathan 00:00:11.080 --> 00:00:14.900 Seyfried. I'm a PhD candidate in history at the 00:00:14.900 --> 00:00:18.839 University of New Mexico, and this is the Historian's 00:00:18.839 --> 00:00:22.739 Notebook, a podcast about how history gets made. 00:00:23.359 --> 00:00:27.480 Season one is titled Molt Cara Companyona. We're 00:00:27.480 --> 00:00:30.539 looking at a document from each day of the first 00:00:30.539 --> 00:00:35.219 year of the reign of King Joan I of Aragon and 00:00:35.219 --> 00:00:38.859 Queen Violant de Bar. Let's begin today's 00:00:38.859 --> 00:01:05.760 document. Today we have another document from 00:01:05.760 --> 00:01:10.159 Violant. It's from her more businessy register, 00:01:10.579 --> 00:01:15.400 the Commune et Curie. And before we get into 00:01:15.400 --> 00:01:18.659 it, it's got a lot of interesting aspects. I 00:01:18.659 --> 00:01:25.140 do want to touch base about the AI usage that 00:01:25.140 --> 00:01:30.290 I'm experimenting with. using OpenClaw and having 00:01:30.290 --> 00:01:36.849 an AI agent do some of the uploading and transferring 00:01:36.849 --> 00:01:41.370 of transcriptions from Gemini to Claude. So it's 00:01:41.370 --> 00:01:46.629 proceeding along. It's got its downsides, that's 00:01:46.629 --> 00:01:50.510 for sure. So let me explain. Well, it's got its 00:01:50.510 --> 00:01:55.760 tremendous upsides. So I tested it out. for one 00:01:55.760 --> 00:02:00.219 register and had it run overnight and it went 00:02:00.219 --> 00:02:04.319 really well in the sense that I was able to get 00:02:04.319 --> 00:02:08.699 some output that I can search and I did a couple 00:02:08.699 --> 00:02:13.819 full text searches and I was able to find some 00:02:13.819 --> 00:02:19.189 interesting things. So the problem is that the 00:02:19.189 --> 00:02:22.550 Gemini part of the process just didn't happen. 00:02:23.229 --> 00:02:27.969 So there's some kinds of restrictions on usage 00:02:27.969 --> 00:02:31.789 that this sort of thing really runs up against 00:02:31.789 --> 00:02:36.090 because what I'm asking to have done here is 00:02:36.090 --> 00:02:41.289 a pretty intensive task, a lot of tokens as they 00:02:41.289 --> 00:02:46.389 say, because the AI Well, the large language 00:02:46.389 --> 00:02:51.330 model has to analyze the image file, find the 00:02:51.330 --> 00:02:57.129 letters, do the paleography, produce a transcription. 00:02:57.830 --> 00:03:02.370 That is asking a lot. It's not the same as having 00:03:02.370 --> 00:03:06.330 an AI agent just manage your calendar for you 00:03:06.330 --> 00:03:09.430 or send some emails. Which, by the way, I would 00:03:09.430 --> 00:03:14.319 not advise doing at this point. um because i 00:03:14.319 --> 00:03:16.319 don't know i just think you should sort of know 00:03:16.319 --> 00:03:18.840 what's going on i don't know maybe i'm just not 00:03:18.840 --> 00:03:20.479 thinking of that in the right way and someone 00:03:20.479 --> 00:03:23.939 can help me understand why that would be a good 00:03:23.939 --> 00:03:29.879 thing to do anyway so i'm really hitting usage 00:03:29.879 --> 00:03:34.539 limits a whole lot and in the second register 00:03:34.539 --> 00:03:38.479 that i'm trying this out with i have to essentially 00:03:38.479 --> 00:03:46.039 have the AI agent the open claw on this machine 00:03:46.039 --> 00:03:51.599 that I'm using delay its requests to Gemini, 00:03:51.740 --> 00:03:55.500 so Basically it has to and I'm not the one that 00:03:55.500 --> 00:04:01.300 thought of that so the open claw agent checked 00:04:01.300 --> 00:04:06.300 what was going on with Gemini and found that 00:04:06.300 --> 00:04:12.210 even though I have signed up for the level of 00:04:12.210 --> 00:04:16.290 service that you pay for, that there's just these 00:04:16.290 --> 00:04:19.350 universal usage limits that this kind of task 00:04:19.350 --> 00:04:24.089 makes it so that you can't do it at the speed 00:04:24.089 --> 00:04:29.850 that it can be done from the agent's point of 00:04:29.850 --> 00:04:35.350 view. So the agent is changing its workflow so 00:04:35.350 --> 00:04:40.240 that instead of sending things to Gemini at the 00:04:40.240 --> 00:04:44.639 speed that it can, it's basically adding in a 00:04:44.639 --> 00:04:50.980 30 second pause between every single document. 00:04:51.839 --> 00:04:57.639 And then that will hopefully comply with the 00:04:57.639 --> 00:05:02.399 usage requirements that Gemini has. I feel like 00:05:02.399 --> 00:05:05.939 in five years time, people will look back at 00:05:05.939 --> 00:05:10.540 this and just kind of chuckle. Well, if there's 00:05:10.540 --> 00:05:14.000 not an AI apocalypse between now and then. But 00:05:14.000 --> 00:05:18.319 if we're still around in five years time, people 00:05:18.319 --> 00:05:21.240 will look back at this and chuckle because these 00:05:21.240 --> 00:05:27.509 usage limits are of course... going to disappear 00:05:27.509 --> 00:05:31.370 as capabilities increase over the next couple 00:05:31.370 --> 00:05:34.350 years. I would be quite shocked if this sort 00:05:34.350 --> 00:05:39.209 of problem would occur after five years of further 00:05:39.209 --> 00:05:43.569 development. Alright, so that's one issue where 00:05:43.569 --> 00:05:49.709 I have not yet successfully replicated my exact 00:05:49.709 --> 00:05:55.069 workflow with open -claw running things because 00:05:55.069 --> 00:05:59.850 of that Gemini usage limit. But Claude is able 00:05:59.850 --> 00:06:04.649 to move at that speed. It is just costly. So 00:06:04.649 --> 00:06:09.189 basically the price that I paid for the first 00:06:09.189 --> 00:06:13.250 register that I did this with from start to end, 00:06:13.310 --> 00:06:17.740 it was about $50. so like that's not going to 00:06:17.740 --> 00:06:21.560 be happening for all of the registers that contain 00:06:21.560 --> 00:06:25.220 content from 1387. I think I'm just going to 00:06:25.220 --> 00:06:30.300 be picking five registers, and still that's a 00:06:30.300 --> 00:06:34.540 chunk of change right there. So the cost barrier 00:06:34.540 --> 00:06:38.360 is significant right now, and I have to make 00:06:38.360 --> 00:06:41.000 some choices because I don't have institutional 00:06:41.000 --> 00:06:44.860 funding for this, and I guess that's something 00:06:44.860 --> 00:06:47.500 I should look into, but that takes a lot of time 00:06:47.500 --> 00:06:52.639 to figure out. All right, well anyway. So those 00:06:52.639 --> 00:06:57.300 are the limitations, the usage limits, and the 00:06:57.300 --> 00:07:02.060 time. So we're going to see what happens with 00:07:02.060 --> 00:07:08.100 this next round when we have OpenClaw building 00:07:08.100 --> 00:07:14.060 in a 30 second pause in the way that it interacts 00:07:14.060 --> 00:07:18.839 with Gemini, it'll probably help to manage things 00:07:18.839 --> 00:07:22.920 with Claude also. But let me talk about the quality 00:07:22.920 --> 00:07:27.500 of the output for a moment, because, well, for 00:07:27.500 --> 00:07:32.259 today's document, I did what I have been doing 00:07:32.259 --> 00:07:36.350 for the last couple weeks, which is I've uploaded 00:07:36.350 --> 00:07:40.050 it to Gemini, had Gemini do a transcription, 00:07:40.790 --> 00:07:46.889 then I upload the image of the document to Claude, 00:07:46.930 --> 00:07:49.829 and I have Claude do a transcription, and then 00:07:49.829 --> 00:07:56.290 I take what Gemini did and I feed that into the 00:07:56.290 --> 00:08:01.790 Claude chat and say reconcile these. Now today's 00:08:01.790 --> 00:08:05.829 document is in Latin. and it's a little bit more 00:08:05.829 --> 00:08:07.829 challenging because, well, I don't know, sometimes 00:08:07.829 --> 00:08:10.689 I get a feeling like the language difference 00:08:10.689 --> 00:08:13.709 doesn't really matter to them too much, but maybe 00:08:13.709 --> 00:08:15.810 the paleography in this one is a little bit more 00:08:15.810 --> 00:08:17.449 difficult. It doesn't, as I'm looking at it, 00:08:17.529 --> 00:08:20.470 it doesn't strike me. I guess there's more abbreviations 00:08:20.470 --> 00:08:23.449 when they're writing in Latin, actually. More 00:08:23.449 --> 00:08:27.050 of a diversity of abbreviation in the Latin documents, 00:08:27.089 --> 00:08:30.350 I think I'm noticing. So I think that's part 00:08:30.350 --> 00:08:34.019 of the issue. and possibly this particular scribe 00:08:34.019 --> 00:08:37.600 is challenging. He's not good old Bartolomeo 00:08:37.600 --> 00:08:41.360 Sirvent. Oh, Bartolomeo Sirvent is my favorite 00:08:41.360 --> 00:08:47.340 scribe, I think. No, this is Andres Granelli, 00:08:47.539 --> 00:08:52.080 who we've had a couple times before. So both 00:08:52.080 --> 00:08:56.419 Gemini and Claude struggled mightily with this 00:08:56.419 --> 00:09:01.159 one, and I actually thought that the product 00:09:01.159 --> 00:09:04.899 that I got after Claude reconciled what Gemini 00:09:04.899 --> 00:09:07.899 had done was so problematic and the document 00:09:07.899 --> 00:09:10.559 was short enough that I just went in and I typed 00:09:10.559 --> 00:09:14.379 in my own transcription. And I mean, I'm not 00:09:14.379 --> 00:09:16.799 good enough to have found everything perfectly, 00:09:16.940 --> 00:09:21.360 but I did enough to be able to make some breakthroughs 00:09:21.360 --> 00:09:24.620 on certain words that had been stumping everybody. 00:09:24.940 --> 00:09:28.750 Like in the fourth line, there's the second word 00:09:28.750 --> 00:09:34.049 it's got a pro abbreviation and then a little 00:09:34.049 --> 00:09:37.529 like looks at first you look at it and you think 00:09:37.529 --> 00:09:42.549 it might be a v and then you realize oh that's 00:09:42.549 --> 00:09:46.309 an x but i didn't quite get to the x i really 00:09:46.309 --> 00:09:48.429 i feel now that i'm looking at i'm like come 00:09:48.429 --> 00:09:51.389 on i really should have understood that that 00:09:51.389 --> 00:09:55.690 was an x and for some reason i was just not wanting 00:09:56.159 --> 00:09:59.899 To go that route when I was looking at it before 00:09:59.899 --> 00:10:02.120 I can't quite figure out what I was thinking 00:10:02.120 --> 00:10:08.019 because it is Pro and then X with a macron and 00:10:08.019 --> 00:10:12.639 above which would make sense to expand is Proxima 00:10:12.639 --> 00:10:20.080 so that line Doesn't seem to have anything about 00:10:20.080 --> 00:10:25.190 time but if you go back to the end of line three 00:10:25.190 --> 00:10:29.409 you've got the word dia day so something like 00:10:29.409 --> 00:10:34.029 the sixth day after and then some kind of form 00:10:34.029 --> 00:10:41.490 of domina or dominica proxima or it could actually 00:10:41.490 --> 00:10:46.429 be the sunday the sunday because what we have 00:10:46.549 --> 00:10:50.629 that would be interesting if that if that first 00:10:50.629 --> 00:10:55.690 word of line four is actually the word for sunday 00:10:55.690 --> 00:11:00.190 like domingue so uh that's super interesting 00:11:00.190 --> 00:11:03.169 i think it is now that i'm working it out because 00:11:03.169 --> 00:11:08.389 the fourth word of line four is quasimodo and 00:11:08.389 --> 00:11:11.769 when i hear quasimodo i immediately think of 00:11:11.769 --> 00:11:15.129 the hunchback of notredame and like the disney 00:11:15.129 --> 00:11:18.720 version that I saw when I was younger, but of 00:11:18.720 --> 00:11:24.059 course it's based on a novel by Victor Hugo. 00:11:24.759 --> 00:11:28.500 And there's a character named Quasimodo in there. 00:11:29.519 --> 00:11:32.620 I think in both of those I have not read the 00:11:32.620 --> 00:11:37.000 novel. Anyway, that is very misleading because 00:11:37.000 --> 00:11:39.460 this is not about the Hunchback of Notre Dame. 00:11:40.360 --> 00:11:45.720 Actually, what I learned today because of Well, 00:11:45.940 --> 00:11:48.100 I think also Gemini picked up on this. I can't 00:11:48.100 --> 00:11:50.740 quite remember, but at least one of the chatbots 00:11:50.740 --> 00:11:55.580 understood that Quasimodo was potentially actually 00:11:55.580 --> 00:12:01.320 the name for a holiday. The Sunday following 00:12:01.320 --> 00:12:05.120 Easter Sunday, one of its names is Quasimodo. 00:12:05.840 --> 00:12:12.659 And that's because the word Quasimodo means something 00:12:12.659 --> 00:12:15.980 like that's like newly changed. And there's something 00:12:15.980 --> 00:12:19.379 about like the baby Jesus newly changed. I don't 00:12:19.379 --> 00:12:21.059 know. I didn't get into it too much. I'll put 00:12:21.059 --> 00:12:24.179 in the Wikipedia page. I'm being super lazy about 00:12:24.179 --> 00:12:30.039 that. All right. So that is just one example 00:12:30.039 --> 00:12:35.899 of how I was able with my intervention to get 00:12:35.899 --> 00:12:39.679 past something that Well, and the chatbots helped. 00:12:39.700 --> 00:12:41.419 We worked together. We worked together. It was 00:12:41.419 --> 00:12:45.259 a real, you know, Ethan Mollick has his book about 00:12:45.259 --> 00:12:47.360 AI, which I highly recommend. I'll put in the 00:12:47.360 --> 00:12:51.639 bibliography for this page. The title is Co -Intelligence, 00:12:51.740 --> 00:12:56.480 and this is absolutely an instance of a co -intelligence. 00:12:57.659 --> 00:13:02.799 So that's the update with AI. And if you haven't 00:13:02.799 --> 00:13:06.799 noticed already, this podcast is almost like 00:13:07.080 --> 00:13:11.980 has like a shadow podcast in it and You know 00:13:11.980 --> 00:13:16.120 the the you know the shadow podcast is a journal 00:13:16.120 --> 00:13:20.779 of how AI has developed in the months that I'm 00:13:20.779 --> 00:13:25.120 looking at these documents and we're adding to 00:13:25.120 --> 00:13:29.879 that as We try our AI agent and we'll see what 00:13:29.879 --> 00:13:36.340 happens but all of that is to say that even when 00:13:36.340 --> 00:13:42.159 the AI agent is able to run an entire register 00:13:42.159 --> 00:13:47.259 of images through Gemini and Claude and to replicate 00:13:47.259 --> 00:13:50.500 what I do, it's still missing that step that 00:13:50.500 --> 00:13:55.100 I used today, which is to notice that the transcription, 00:13:55.240 --> 00:13:58.399 for whatever reason, was particularly low quality 00:13:58.399 --> 00:14:02.169 and to intervene before... the final thing. I 00:14:02.169 --> 00:14:04.250 can still intervene as I'm looking at these things 00:14:04.250 --> 00:14:06.629 later, but just think about the implications 00:14:06.629 --> 00:14:09.309 for what I'm trying to do, something like a full 00:14:09.309 --> 00:14:12.450 text search for letters about certain things. 00:14:12.970 --> 00:14:15.549 So it's just a reminder that if I'm going to 00:14:15.549 --> 00:14:19.509 be searching the big data that comes out of what 00:14:19.509 --> 00:14:25.159 the AI agent, what the AI agent generates, I 00:14:25.159 --> 00:14:27.659 need to be searching for different things in 00:14:27.659 --> 00:14:29.799 different ways and not just one little search 00:14:29.799 --> 00:14:32.039 and saying, oh, nothing came up, that must be, 00:14:32.059 --> 00:14:36.559 and it's not there. Nope, nope, nope. It's gotta 00:14:36.559 --> 00:14:39.840 be treated as like a rudimentary data set and 00:14:39.840 --> 00:14:43.720 not anything like a critical edition. All right, 00:14:43.720 --> 00:14:47.840 let's talk about this document. This document 00:14:47.840 --> 00:14:53.620 is an example of interfaith context. So Violant 00:14:53.620 --> 00:14:57.940 is telling the addressee, who appears right in 00:14:57.940 --> 00:15:02.799 that first line, Cresques Bonafos. Crescques is 00:15:02.799 --> 00:15:05.519 a pretty common Jewish name, and in fact, that 00:15:05.519 --> 00:15:10.440 name came up. I don't know if it's a family relation 00:15:10.440 --> 00:15:14.899 or just a coincidence, but in the letter from 00:15:14.899 --> 00:15:20.379 Joan about the map, the Catalan Atlas, the Majorcan 00:15:20.379 --> 00:15:26.799 school of cartographers had the Cresques family, 00:15:26.919 --> 00:15:30.019 the father, son, and probably others who were 00:15:30.019 --> 00:15:33.200 cartographers there. But I don't see that there's 00:15:33.200 --> 00:15:36.700 any connection to this particular Cresques in 00:15:36.700 --> 00:15:40.799 today's document, because this person lives in 00:15:40.799 --> 00:15:45.220 Figueres. Now I'm pretty sure we're talking about 00:15:45.220 --> 00:15:51.080 Figueres. The third from the last word of line 00:15:51.080 --> 00:15:55.360 one has the two Fs, which clues you in that it's 00:15:55.360 --> 00:16:01.039 a proper name, and it's got F -I -G -I -I -S 00:16:01.039 --> 00:16:06.000 as the spelling, and that diagonal stroke above 00:16:06.000 --> 00:16:08.919 the word I think is like the dot on the I. I 00:16:08.919 --> 00:16:11.899 don't think that there's an abbreviation. I think 00:16:11.899 --> 00:16:14.600 that the Latin name for Figueres just doesn't 00:16:14.600 --> 00:16:20.639 have the R. but I'm not entirely sure. The chatbots, 00:16:20.639 --> 00:16:23.059 that's what they came up with too, that it's 00:16:23.059 --> 00:16:25.559 Figueres. We're going to go with that for now, 00:16:25.559 --> 00:16:30.519 but I still want to say that's provisional. And 00:16:30.519 --> 00:16:39.919 the word before de Figueres or de Figuis is Iudei, 00:16:40.419 --> 00:16:45.919 actually. So he's being identified as a Jew of 00:16:45.960 --> 00:16:50.879 the town of Figueres. And then what's really 00:16:50.879 --> 00:16:54.620 interesting is that really quick into it on the 00:16:54.620 --> 00:16:58.299 second line we see that there's been a supplication, 00:16:58.700 --> 00:17:04.579 some kind of petition, and the fourth word is 00:17:04.579 --> 00:17:07.839 regino. I think that means that it was a supplication 00:17:07.839 --> 00:17:15.200 to the queen. And then we have UX and then t 00:17:15.200 --> 00:17:21.619 -u -e. And the macron over the u -x is very faint, 00:17:21.680 --> 00:17:26.940 but it is there. So this is uxore tuae, your 00:17:26.940 --> 00:17:30.599 wife. And then we've got some adjectives afterwards, 00:17:31.319 --> 00:17:37.180 paupis et miserabilis. So she's poor and miserable. 00:17:37.740 --> 00:17:41.420 So something has gone wrong with this marriage 00:17:41.420 --> 00:17:47.099 because the wife of Cresques Bonifos has made 00:17:47.099 --> 00:17:52.880 some kind of supplication to the Queen. And the 00:17:52.880 --> 00:17:55.440 chatbots were thinking that the name of the wife 00:17:55.440 --> 00:17:59.160 is Regina. I do not think so. I don't think so. 00:17:59.480 --> 00:18:04.920 I think that's dative, the Regino, and that it 00:18:04.920 --> 00:18:10.619 means that the supplication was for or to the 00:18:10.619 --> 00:18:15.480 Queen. And so basically the letter is mandating 00:18:15.480 --> 00:18:21.940 that Crescas Boniface on the sixth day after 00:18:21.940 --> 00:18:28.380 Quasimodo shows up to Barcelona and answers this. 00:18:28.460 --> 00:18:31.079 And they're gonna work it out. And they're gonna 00:18:31.079 --> 00:18:35.170 get some justice. And I think that's... pretty 00:18:35.170 --> 00:18:37.390 much what it amounts to. There's some other intriguing 00:18:37.390 --> 00:18:40.589 words in this letter that I'm quite, I'm not 00:18:40.589 --> 00:18:44.589 quite sure how it all fits together. And I think 00:18:44.589 --> 00:18:47.670 the paleography prevented the chat bots from 00:18:47.670 --> 00:18:49.789 getting like a really great translation of this 00:18:49.789 --> 00:18:53.950 into English. But I think that this is absolutely 00:18:53.950 --> 00:19:00.769 fascinating that we have the queen getting involved 00:19:00.769 --> 00:19:06.359 in a complaint. about a husband being in that 00:19:06.359 --> 00:19:10.240 complaint has been submitted by the wife and 00:19:10.240 --> 00:19:14.960 that this is a Jewish couple. And so I think 00:19:14.960 --> 00:19:19.500 this is super interesting. There's another instance 00:19:19.500 --> 00:19:22.920 that has come up in the historiography that this 00:19:22.920 --> 00:19:26.660 reminds me of, and it's in the latest book by 00:19:26.660 --> 00:19:32.329 Marie Kelleher, The Hungry City. and Marie Kelleher 00:19:32.329 --> 00:19:38.269 tells the story of Bonadona, who was also a Jewish 00:19:38.269 --> 00:19:43.250 woman who used the legal system to get justice 00:19:43.250 --> 00:19:48.029 against a husband. And so I think this is a super 00:19:48.029 --> 00:19:51.450 interesting echo of that because the episode 00:19:51.450 --> 00:19:54.710 that Marie Kelleher talks about occurred earlier 00:19:54.710 --> 00:19:58.289 in the century in a different a different monarch. 00:19:59.269 --> 00:20:00.809 And well, I don't even know if that one went 00:20:00.809 --> 00:20:05.630 to as high as the king or queen, actually. So 00:20:05.630 --> 00:20:08.569 here, I mean, they're going all the way to the 00:20:08.569 --> 00:20:14.349 chief legal authority, since this is in Violant's 00:20:14.349 --> 00:20:17.470 portfolio. So it's super interesting, and I really 00:20:17.470 --> 00:20:22.009 want to know what happened here. And this is 00:20:22.009 --> 00:20:26.670 where, you know, my progress using the AI agent 00:20:26.880 --> 00:20:30.859 can come in, because once this is improved, and 00:20:30.859 --> 00:20:36.000 once the cost comes down, there's so many records 00:20:36.000 --> 00:20:38.700 in the Archive of the Crown of Aragon, not just 00:20:38.700 --> 00:20:42.259 the letters from the royal family, but all kinds 00:20:42.259 --> 00:20:45.759 of civic records, and then there's other in the 00:20:45.759 --> 00:20:49.150 same kind of paleography, there's the same kind 00:20:49.150 --> 00:20:52.049 of script, there's other records from other archives, 00:20:52.049 --> 00:20:55.549 and to make them full -text searchable, one day, 00:20:55.890 --> 00:20:58.789 maybe not too far in the future, I would be able 00:20:58.789 --> 00:21:02.990 to go in and do a search for Cresques Boniface 00:21:02.990 --> 00:21:08.329 and find out more about what was happening here, 00:21:08.329 --> 00:21:11.690 because in its usual form we have a letter that 00:21:11.690 --> 00:21:13.789 doesn't really state specific things. It's all 00:21:13.789 --> 00:21:19.099 just, you know, the case. or the complaint, right, 00:21:19.240 --> 00:21:24.079 or the supplication. All right, thanks for listening 00:21:24.079 --> 00:21:27.059 to this episode of the Historian's Notebook, 00:21:27.339 --> 00:21:31.480 season one. Molt cara companyona. If you are 00:21:31.480 --> 00:21:33.519 leaving with more questions than you arrived 00:21:33.519 --> 00:21:38.119 with, I've done my job. Remember, the motto of 00:21:38.119 --> 00:21:42.660 the Historian's Notebook is dissatisfaction guaranteed. 00:21:43.400 --> 00:21:46.259 Visit the website. to see an image of today's 00:21:46.259 --> 00:21:51.759 document and additional show notes. The AI chat 00:21:51.759 --> 00:21:54.220 transcripts for today's episode I think are going 00:21:54.220 --> 00:21:56.380 to be really interesting to look at, the one 00:21:56.380 --> 00:21:59.839 from Claude in particular. There will be two 00:21:59.839 --> 00:22:02.640 that you can look at. And listen again tomorrow 00:22:02.640 --> 00:22:05.680 to hear about the next day in the first year 00:22:05.680 --> 00:22:09.940 of the reign of King Joan I of Aragon and Queen 00:22:09.940 --> 00:22:13.799 Violant de Bar. In the meantime, take care.