I started a quick transcription here -- not enough time to complete more than half the first column, but some scans and very rough OCR are here if anyone is interested in contributing:
EDIT: If you have access to a multi-modal LLM, the rough transcription + the column scan and the instruction to "OCR this text, keep linebreaks" gives a _very good_ result.
EDIT 2: Rough draft, needs some proofreading and corrections:
Seems like you don't need an LLM, you just need a human who (1) likes reading Stoker and (2) touch-types. :) I'd volunteer, if I didn't think I'd be duplicating effort at this point.
(I've transcribed various things over the years, including Sonia Greene's Alcestis [1] and Holtzman & Kershenblatt's "Castlequest" source code [2], so I know it doesn't take much except quick fingers and sufficient motivation. :))
I finished a very rough, tesseract + LLM transcription, but it absolutely needs editing passes.
I've done transcription in the past myself (did two books for standard ebooks with some from-scratch transcription and lots of editing) and I know the pain. I've always found it easier to fix up OCR than type the whole thing by hand because I've found my error rate of eyeball transcription to be higher.
If you want to tackle the proofing passes, I'm happy to add you to the repo :)
In the scan, where it says "and shortly came to the edge of the Punchbowl and easted my eyes on its beauty" OP changed "easted" to "cast" and the tumbler one says "easted[sic]" ([sic] is theirs). I wonder if it's supposed to be "feasted".
I remember reading somewhere- I think it was in an annotated addition of Dracula, or maybe it was a journal article- that said that Bram Stoker wrote a large number of novels but everything he wrote other than Dracula was awful. Per Wikipedia he wrote 14 books, supposedly he was only able to write one good one.
For me it feels like Stokers dracula is only so popular because it's where all the tropes come from, not because it's particularly well written, or something like that.
It's one of those firsts that established a genre.
I know Stoker didn't invent vampires, but they came into western English speaking culture through his Dracula.
I am not a literary critic, but I very much enjoyed Dracula. When I read it, I did not know there were claims he wasn't a good writer, so I had no bias, I simply liked it quite a bit.
I don't mean to disparage this particular instance at all, as it seems pretty great. But I wonder if the rise of llms is going to make scams that sounds a lot like this much easier in the future. I think at the moment it's hard to make something really sound like a particular author without a lot of work, but that will probably change in the future.
Sure, people can do scams but it will be way more interesting to apply them to finding stuff like this. Up through now, literary treasures and open secrets are sitting out waiting to be recognized.
And why bother with trying to deceive when one could build reputation for creating truly good fan fiction based off real source material.
Just because tech can be used to abuse trust doesn't mean it will be the most interesting and commonly recognized thing to do with it.
I’ve found that it’s not uncommon for an interested individual to find details that have not been documented or “found” by others. I collect video games and have found variants of popular games that have been otherwise undocumented on any list or archive that I was aware of. I’ve found audio recordings from the 90s that seemingly have no recorded history on the internet.
These aren’t things historians have had hundreds of years to document, but several thousand or more people have been on this space long before I was looking at it more intently than I could ever and I still come across things from time to time that weren’t known to exist.
Likewise, in the past month I’ve spent an unfortunate amount of time reading laws and board bylaws and it doesn’t take long to find long forgotten rules that are being actively violated. Even outside of code, documentation is hard.
Tyler Cowen recently interviewed a historian (Alan Taylor), and they approached this subject near the end of the episode—how much the job of a historian still involves browsing undigitized material sitting on a shelf in a cold room somewhere. Around 3215 seconds* in:
> And then there's also a kind of notion that everything is there online when in point of fact lots of information about the past still only exists in archives
* of the audio version, that is; at that timestamp in the YouTube video, they're discussing the question "How will large language models change historical research"—interviewee's response: he doesn't know
This happens often when going down the rabbit hole on a niche project. For example, repair and restoration of Persian rugs.
There are many details to the craft that are hinted at in variety of formats, (youtube videos, blog entries, etc) but the clear truths are not clearly stated anywhere. These are stored in the minds and practices of artisans.
Try skimming the Wikipedia articles on some major authors of that era, to get a sense for how much short (or serialized) fiction & poetry was routinely published in newspapers and magazines back then.
Without some specific clues, a real historian would not be looking for Bram Stoker stories in an 1890 issue of the Daily Express Dublin Edition. He'd be skimming through the archives of many of the newspapers & magazines published in an era and geographic region, cataloging authors & stories & poems. "Success" would be just compiling a well-done catalog. 15 minutes of fame in the popular press could equally well result from finding some unknown early work by James Joyce, or Winston Churchill, or George Bernard Shaw, or Oscar Wilde, or Yeats, or ...
How would copyright law apply here? Would this fall into the public domain immediately? I read that Irish law is that it would be "70 years from date first made available to the public". Since published in a newspaper, I would assume this would be public domain now. Correct?
If this was an unpublished manuscript, rights of first publication would apply and it might be covered by a kind of copyright that would vary depending on the country. Since this was "rediscovered" after first being unambiguously published back in the 1890s, it's pretty clearly in the public domain.
OP got incredibly lucky though that the author's name was included in the original publication - things like this (i.e. contributions to newspapers or magazines) were often published under obscure pseudonyms, initials, puzzling hints like "By the author of Such-and-such" or no author indication at all.
I _think_ UK copyright law would matter here, since at the time the story was published (1890) the Ireland was part of the UK (Ireland gained independence in 1921.)
If UK copyright applied, then the story would have entered public domain in 1932. The term of copyright for published works at the time as 7 years after the authors death, or 42 years, whichever was longer.
Seems like a non pessimistic idea of something LLMs could help us out with. Mass analysis of old texts for new finds like this. If this one exists surely there are many more just a mass analysis away
I accidentally got Zed to parse way more code than I intended last night and it cost close to $2 on the anthropic API. All I can think is how incredibly expensive it would be to feed an LLM text in hopes of making those connections. I don’t think you’re wrong, though. This is the territory where their ability to find patterns can feel pretty magical. It would cost many, many, many $2 though
> I accidentally got Zed to parse way more code than I intended last night and it cost close to $2 on the anthropic API
Is that one API call or some out of control process slinging 100s of requests?
Must have been a ton of data, as their most expensive model (Opus) seems to $15 per million input tokens. I guess if you just set it to use an entire project as the input, you'll hit 1m input tokens quickly.
Come to think of it, I’m not sure how Zed performs LLM requests with the inline assistant.
I wasn’t working in an enormous file, but I meant to highlight a block and accidentally highlighted the entire file and asked it to do something that made no sense in that context. It did its best to do something with the situation and eventually ran out of steam, haha. It’s possible that multiple requests needed to be made, or I was around the 200k context window.
Previous to this I’m fairly sure most of my requests cost fractions of pennies. My credit takes ages to decrease by any meaningful amount. Except until last night. It’s normally an extremely cost-effective tool for me.
This is a pretty good case for just using a local model. Even if it's 50% worse than Anthropic or whatever the gap is now between open models and proprietary state of the art, it's still likely 'good enough' to categorize a story in an old newspaper as missing from an author's known bibliography.
Good point. I use llama3.1 for a lot of small tasks and rarely feel like I need to use Claude instead. It’s fine. I’m even running the model a (big) step down from 70b, because I’ve only got 32GB of ram. It’s a solid model that probably costs me next to nothing to run.
Why? Old texts would be out of copyright, and even if they weren't, as long as you're not publishing the source material or anything containing the source material (or anything that can verbatim output the source), it seems you'd be in the clear.
I’m concerned things like this will just be gone forever in the digital era. Paper and film are great storage mediums. I know this was on a screen but would it have still existed if it wasn’t on paper first?
Hard disks are great storage mediums when we don't purposely set fire to them to preserve the profits of large corporations. The Internet Archive is perfectly capable of preserving things, unless copyright holders manage to shut them down for short-term profit.
IA shouldn't try to wage war against copyright. They should leave that to other entities.
IA should be an archivist organization first and foremost and abandon the idea of making books, movies, and music publicly available. That's just painting a target on their back and risking their goal of preserving a snapshot of our time.
The wayback machine is great, though, and they should keep doing that.
Agreed, and I think it's important to note that paper doesn't have any sort of DRM encumbrance on it. I seriously think that at some point in the next few decades, the "pirates" who right now are hated and prosecuted vigorously by all the "rightsholders" may turn out to be venerable heroes for having preserved the creations.
Imagine if we had found Bram Stokers work, and it was also encrypted mumbo jumbo that is now useless to us. We'll likely never know what we lost.
Thank you! I've been using this word in portuguese (amador) and its so _so_ clear in that language, even so, I hadn't realized. Amar -> Amador (the one who loves it). Quite clearly.
The point is that "amateur" means literally "lover" in latin. While "professional" means "for money" today, in latin it meant "to profess a vow to do it with high standards".
For example, you can be a professional, but do things "pro bono" (for free or for public good) or "pro lucro" (for money).
"Doing something was a high standard" is still the main meaning of the word "professionell" in German. So someone can make something "unprofessionell" for money or "professionell" without payment.
Another word of classical origin with a striking difference is the meaning of the word "pathetisch" in German, which means "(exaggeratedly) passionate", which corresponds more or less to the meaning of the Ancient Greek word "pathetikos".
Dilletante already exists to mean someone who doesn’t do something with seriousness and amateur doesn’t carry the same connotation as amateurish anyway so you don’t really need a new word.
The term 'nerd' needs to complete its rehabilitation like 'geek' has the last 20 years. It's the most concise term I can think of when describing someone who is enthusiastic, focused, and knowledgable on a subject. I think it's a badge of honor
Yeah but it's often intended as an insult. Especially as the adjective amateurish, or phrases like the work of an amateur.
Amateur historian could never be an insult, because it's actually better to have a real career in something substantial, and do the history stuff on the side as a hobby.
I don't know why people get obsessed over things like this. Finding significance in something because it's written by an entity whose name is popular makes no sense.
I started a quick transcription here -- not enough time to complete more than half the first column, but some scans and very rough OCR are here if anyone is interested in contributing:
https://github.com/mmastrac/gibbet-hill
Top and bottom halves of the page in the repo here:
https://github.com/mmastrac/gibbet-hill/blob/main/scan-1.png https://github.com/mmastrac/gibbet-hill/blob/main/scan-2.png
EDIT: If you have access to a multi-modal LLM, the rough transcription + the column scan and the instruction to "OCR this text, keep linebreaks" gives a _very good_ result.
EDIT 2: Rough draft, needs some proofreading and corrections:
https://github.com/mmastrac/gibbet-hill/blob/main/story.md
Seems like you don't need an LLM, you just need a human who (1) likes reading Stoker and (2) touch-types. :) I'd volunteer, if I didn't think I'd be duplicating effort at this point.
(I've transcribed various things over the years, including Sonia Greene's Alcestis [1] and Holtzman & Kershenblatt's "Castlequest" source code [2], so I know it doesn't take much except quick fingers and sufficient motivation. :))
[1] https://quuxplusone.github.io/blog/2022/10/22/alcestis/
[2] https://quuxplusone.github.io/blog/2021/03/09/castlequest/
EDIT: ...and as I was writing that, you seem to have finished your transcription. :)
I finished a very rough, tesseract + LLM transcription, but it absolutely needs editing passes.
I've done transcription in the past myself (did two books for standard ebooks with some from-scratch transcription and lots of editing) and I know the pain. I've always found it easier to fix up OCR than type the whole thing by hand because I've found my error rate of eyeball transcription to be higher.
If you want to tackle the proofing passes, I'm happy to add you to the repo :)
Use LiveText API. Much much better accuracy than Tesseract. You can rent access to it.
Too late. You have already been scooped by, of course, tumblr:
<https://woodsfae.tumblr.com/post/764918993659330560/gibbet-h...>
In the scan, where it says "and shortly came to the edge of the Punchbowl and easted my eyes on its beauty" OP changed "easted" to "cast" and the tumbler one says "easted[sic]" ([sic] is theirs). I wonder if it's supposed to be "feasted".
A battle of a Tumblr user named Woodsfae versus advanced LLM transcribing new goth literature?
That's like bringing a knife to a gun fight my friend, never underestimate the power of a committed Tumblr user
I tried extracting the content using Google Gemini 1.5 Pro 002 using https://aistudio.google.com/ - the first page (scan-2) worked fantastically well, the second page not so much. Here's what I got so far: https://gist.github.com/simonw/ba87f507ef5c11d3335959c055533...
I cropped the columns out into six files -- it might have an easier time with these:
https://github.com/mmastrac/gibbet-hill/blob/main/col-1-a.pn...
…and my wife's Halloween present has been printed.
Tip: Load the pngs into Preview, hit "Auto Levels," and crank up "Sharpness" on each one. Looks pretty good!
[dead]
probably you would want to get the project gutenberg people onto it
I went ahead and made a post over at the PG proofreaders site (pgdp.net) to make them aware.
I remember reading somewhere- I think it was in an annotated addition of Dracula, or maybe it was a journal article- that said that Bram Stoker wrote a large number of novels but everything he wrote other than Dracula was awful. Per Wikipedia he wrote 14 books, supposedly he was only able to write one good one.
Not a novel, but the short story "Dracula's Guest" I thought was quite good. I was sad it was so short.
I suspect you're getting downvoted by people who haven't actually read anything by Stoker.
My wife has read most of his stuff. I know because I buy it for her. She says aside from Dracula, most of it is not great.
For me it feels like Stokers dracula is only so popular because it's where all the tropes come from, not because it's particularly well written, or something like that.
It's one of those firsts that established a genre.
I know Stoker didn't invent vampires, but they came into western English speaking culture through his Dracula.
I am not a literary critic, but I very much enjoyed Dracula. When I read it, I did not know there were claims he wasn't a good writer, so I had no bias, I simply liked it quite a bit.
You can read it here: https://catalogue.nli.ie/Record/vtls000924296
Go full screen and go to page 2 it starts at about the middle.
Brian Cleary will be discussing his findings next Saturday in Dublin, as part of the Bram Stoker Festival: https://bramstokerfestival.com/en/events/an-extraordinary-br...
I don't mean to disparage this particular instance at all, as it seems pretty great. But I wonder if the rise of llms is going to make scams that sounds a lot like this much easier in the future. I think at the moment it's hard to make something really sound like a particular author without a lot of work, but that will probably change in the future.
Sure, people can do scams but it will be way more interesting to apply them to finding stuff like this. Up through now, literary treasures and open secrets are sitting out waiting to be recognized.
And why bother with trying to deceive when one could build reputation for creating truly good fan fiction based off real source material.
Just because tech can be used to abuse trust doesn't mean it will be the most interesting and commonly recognized thing to do with it.
I can see it now:
"3 million lost works of Shakespeare found"
"Honey, come look! I've found some information all the world's top historians missed."
I’ve found that it’s not uncommon for an interested individual to find details that have not been documented or “found” by others. I collect video games and have found variants of popular games that have been otherwise undocumented on any list or archive that I was aware of. I’ve found audio recordings from the 90s that seemingly have no recorded history on the internet.
These aren’t things historians have had hundreds of years to document, but several thousand or more people have been on this space long before I was looking at it more intently than I could ever and I still come across things from time to time that weren’t known to exist.
Likewise, in the past month I’ve spent an unfortunate amount of time reading laws and board bylaws and it doesn’t take long to find long forgotten rules that are being actively violated. Even outside of code, documentation is hard.
Tyler Cowen recently interviewed a historian (Alan Taylor), and they approached this subject near the end of the episode—how much the job of a historian still involves browsing undigitized material sitting on a shelf in a cold room somewhere. Around 3215 seconds* in:
> And then there's also a kind of notion that everything is there online when in point of fact lots of information about the past still only exists in archives
<https://conversationswithtyler.com/episodes/alan-taylor/>
* of the audio version, that is; at that timestamp in the YouTube video, they're discussing the question "How will large language models change historical research"—interviewee's response: he doesn't know
This happens often when going down the rabbit hole on a niche project. For example, repair and restoration of Persian rugs.
There are many details to the craft that are hinted at in variety of formats, (youtube videos, blog entries, etc) but the clear truths are not clearly stated anywhere. These are stored in the minds and practices of artisans.
"missed" might be taken to imply that one or more of them had ever bothered to look.
Well, even if people were looking, this sort of thing is a lot of right place and right time.
Try skimming the Wikipedia articles on some major authors of that era, to get a sense for how much short (or serialized) fiction & poetry was routinely published in newspapers and magazines back then.
Without some specific clues, a real historian would not be looking for Bram Stoker stories in an 1890 issue of the Daily Express Dublin Edition. He'd be skimming through the archives of many of the newspapers & magazines published in an era and geographic region, cataloging authors & stories & poems. "Success" would be just compiling a well-done catalog. 15 minutes of fame in the popular press could equally well result from finding some unknown early work by James Joyce, or Winston Churchill, or George Bernard Shaw, or Oscar Wilde, or Yeats, or ...
How would copyright law apply here? Would this fall into the public domain immediately? I read that Irish law is that it would be "70 years from date first made available to the public". Since published in a newspaper, I would assume this would be public domain now. Correct?
If this was an unpublished manuscript, rights of first publication would apply and it might be covered by a kind of copyright that would vary depending on the country. Since this was "rediscovered" after first being unambiguously published back in the 1890s, it's pretty clearly in the public domain.
OP got incredibly lucky though that the author's name was included in the original publication - things like this (i.e. contributions to newspapers or magazines) were often published under obscure pseudonyms, initials, puzzling hints like "By the author of Such-and-such" or no author indication at all.
I _think_ UK copyright law would matter here, since at the time the story was published (1890) the Ireland was part of the UK (Ireland gained independence in 1921.)
If UK copyright applied, then the story would have entered public domain in 1932. The term of copyright for published works at the time as 7 years after the authors death, or 42 years, whichever was longer.
Funnily enough there was a reddit post from around the time the manuscript was discovered (but before it was announced) asking a similar question
Yes, it's public domain
Seems like a non pessimistic idea of something LLMs could help us out with. Mass analysis of old texts for new finds like this. If this one exists surely there are many more just a mass analysis away
I accidentally got Zed to parse way more code than I intended last night and it cost close to $2 on the anthropic API. All I can think is how incredibly expensive it would be to feed an LLM text in hopes of making those connections. I don’t think you’re wrong, though. This is the territory where their ability to find patterns can feel pretty magical. It would cost many, many, many $2 though
> I accidentally got Zed to parse way more code than I intended last night and it cost close to $2 on the anthropic API
Is that one API call or some out of control process slinging 100s of requests?
Must have been a ton of data, as their most expensive model (Opus) seems to $15 per million input tokens. I guess if you just set it to use an entire project as the input, you'll hit 1m input tokens quickly.
Come to think of it, I’m not sure how Zed performs LLM requests with the inline assistant.
I wasn’t working in an enormous file, but I meant to highlight a block and accidentally highlighted the entire file and asked it to do something that made no sense in that context. It did its best to do something with the situation and eventually ran out of steam, haha. It’s possible that multiple requests needed to be made, or I was around the 200k context window.
Previous to this I’m fairly sure most of my requests cost fractions of pennies. My credit takes ages to decrease by any meaningful amount. Except until last night. It’s normally an extremely cost-effective tool for me.
This is a pretty good case for just using a local model. Even if it's 50% worse than Anthropic or whatever the gap is now between open models and proprietary state of the art, it's still likely 'good enough' to categorize a story in an old newspaper as missing from an author's known bibliography.
Good point. I use llama3.1 for a lot of small tasks and rarely feel like I need to use Claude instead. It’s fine. I’m even running the model a (big) step down from 70b, because I’ve only got 32GB of ram. It’s a solid model that probably costs me next to nothing to run.
The problem with copyright is going to be a big hurdle though.
Why? Old texts would be out of copyright, and even if they weren't, as long as you're not publishing the source material or anything containing the source material (or anything that can verbatim output the source), it seems you'd be in the clear.
You are right! I forgot about this completely.
If we go to the era of public domain, there is no worry about copyright.
I’m concerned things like this will just be gone forever in the digital era. Paper and film are great storage mediums. I know this was on a screen but would it have still existed if it wasn’t on paper first?
Hard disks are great storage mediums when we don't purposely set fire to them to preserve the profits of large corporations. The Internet Archive is perfectly capable of preserving things, unless copyright holders manage to shut them down for short-term profit.
IA shouldn't try to wage war against copyright. They should leave that to other entities.
IA should be an archivist organization first and foremost and abandon the idea of making books, movies, and music publicly available. That's just painting a target on their back and risking their goal of preserving a snapshot of our time.
The wayback machine is great, though, and they should keep doing that.
What are you referring to here? Hopefully not the secure destruction of hard disks.
The law's preference for 120 years of copyright instead of the preservation of culture. IA should be state-funded.
How does copyright relate to burning hard drives?
Agreed, and I think it's important to note that paper doesn't have any sort of DRM encumbrance on it. I seriously think that at some point in the next few decades, the "pirates" who right now are hated and prosecuted vigorously by all the "rightsholders" may turn out to be venerable heroes for having preserved the creations.
Imagine if we had found Bram Stokers work, and it was also encrypted mumbo jumbo that is now useless to us. We'll likely never know what we lost.
All this, and yet no link to read it?
It's 134 years old but hasn't been published as a book yet, so surely it requires 100 years of copyright protection starting today!
https://news.ycombinator.com/item?id=41905844
Used to be that writers were paid by the word and novels were serialized.
It's funny (ironic?), but when I read "an amateur {insert occupation} has"
I mentally replace "an amateur" with "a talented and passionate"
For me, amateur just doesn't mean the insult that it meant when I was a youngster.
The word 'amateur' originates from the Latin word for 'lover'.
Thank you! I've been using this word in portuguese (amador) and its so _so_ clear in that language, even so, I hadn't realized. Amar -> Amador (the one who loves it). Quite clearly.
Exactly, and "professional" means they do it for money.
The point is that "amateur" means literally "lover" in latin. While "professional" means "for money" today, in latin it meant "to profess a vow to do it with high standards".
For example, you can be a professional, but do things "pro bono" (for free or for public good) or "pro lucro" (for money).
"Doing something was a high standard" is still the main meaning of the word "professionell" in German. So someone can make something "unprofessionell" for money or "professionell" without payment.
Another word of classical origin with a striking difference is the meaning of the word "pathetisch" in German, which means "(exaggeratedly) passionate", which corresponds more or less to the meaning of the Ancient Greek word "pathetikos".
"Vocation" has undergone a similar shift; originally it meant a calling, or a summons.
It still does.
Just don't go looking for the etymology of "vanilla"
But amateur has taken on a negative connotation in the common vernacular.
"Amateurish" or "amateurishly" feel damning and assertions about a certain absence of quality or attention to detail.
Describing someone as a "total amateur" feels a bit like calling them a hack.
This needs a separate word or concept.
Dilletante already exists to mean someone who doesn’t do something with seriousness and amateur doesn’t carry the same connotation as amateurish anyway so you don’t really need a new word.
The term 'nerd' needs to complete its rehabilitation like 'geek' has the last 20 years. It's the most concise term I can think of when describing someone who is enthusiastic, focused, and knowledgable on a subject. I think it's a badge of honor
There's "aficionado", though that feels a little pretentious
We could try reclaiming the word.
I have never thought of it as an insult, just meaning they don't do it for money.
For me, amateur generally just translates as "not paid for his services."
Yeah but it's often intended as an insult. Especially as the adjective amateurish, or phrases like the work of an amateur.
Amateur historian could never be an insult, because it's actually better to have a real career in something substantial, and do the history stuff on the side as a hobby.
Evolving Wikipedia Entry on the Story "Gibbet Hill" [0]. Plot Summary described on the page.
[0] https://en.wikipedia.org/wiki/Gibbet_Hill_(short_story)
Does the name "Bram Stoker" not carry any weight?
Yes, but "Dracula author" carries more, and headlines aim to reach as many people as possible.
For some reason his name is in the page's <head> but not in the article's title.
Pro search tip- its Stoker not Stroker
porn parody potential there
It does here!
Insofar as he's associated with "that Dracula story and movie", yes.
I don't know why people get obsessed over things like this. Finding significance in something because it's written by an entity whose name is popular makes no sense.