- Options
- Login
- Desktop Style
Options
Using AI to do research - Specifically online searching
- Using AI to do research - Specifically online searching
-
RE: Using AI to do research - Specifically online searching
Mark Knowles > 6 hours ago
Of course, there are undoubtedly many ciphers in archives and libraries that are not listed anywhere online and the only way to find such ciphers is to manually search for them. Obviously, identifying which are likely to be the most productive archives or libraries to search and where in those archives or libraries to search is an important part of research.
As I have stated before I am very keen to find examples of Milanese ciphers from between 1425 and 1434 when Francesco Barbavara was Ducal Secretary in Milan.(Although Milanese ciphers from 1435 to 1439 would also be interesting.)
I must admit that I find Family Archives frustrating as they are often not inventorised at all even when they have been donated to a well-known archive(Of course, when they are in private hands there is blackness. I call these the dark archives.) And Family Archives can have all sorts of things in them as the discovery of the Albertoni cipher ledger proves. -
RE: Using AI to do research - Specifically online searching
Mark Knowles > 5 hours ago
(6 hours ago)DG97EEB Wrote: You are not allowed to view links. Register or Login to view.Hi Mark,
I ran a combination GPT and Claude.. Hope it helps
Ed
Thanks for that.
When I look at these kinds of AI researches as with the previous stuff I tried with GPT and Gemini I don't know how to react. My initial reaction is that this is brilliant and very impressive. However when I start to work through the details I am not sure. First, I wonder to what extent it is feeding back to me in a slightly garbled/misinterpreted form what I have posted on Nick Pelling's blog as I have posted a lot about the ciphers that I have located. However, funnily enough I like it's confidence which I find encouraging. I find the reference to "Andrea Barbarigo’s private merchant cipher" intriguing. I have been in touch with Dr. Iordanou before, so I could follow that up. Paolo Bonavoglia says he has searched the Venice archives for diplomatic ciphers of that period thoroughly, so either he hasn't searched it as thoroughly as he thinks or there isn't any more to be found there. It seems to not have picked up on the 1397 Milanese cipher in the Chronicles of Lucca by Giovanni Sercambi as the earliest example of homophones we know of.
It is right that the State Archives of Mantua, Modena and Florence Archives need to be searched from top to bottom as they have already yielded much interesting materials, but there may be more to be found there.
It may be worth confirming that I have seen the following that it mentions:
1404–1418 enciphered correspondence
Jacobo de Faitinellis, Dino ser Paci, Iohannello Thieri, Guido da Pietrasanta, Nicolao da Moncicoli, Nicolao Arnolfini.
Foreign Correspondence, b. 25, sez. 3, Rome
Look for the 1426–1430 enciphered letters: Ricardo de Mutiliana, Pandolfo Malatesta, unsigned Rome letter, Francesco de Catalenis.
---
The tendency of these AI to hallucinate or just misunderstand what was previously been written makes me wonder whether it's references are genuine and correctly dated.I think it would be worth asking the AIs for the precise "shelfmark" for these ciphers and where they saw the references to them.
I like to think that I am not so proud that I care where a new lead comes from and that I will follow it up regardless. -
RE: Using AI to do research - Specifically online searching
DG97EEB > 5 hours ago
(5 hours ago)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.
(6 hours ago)DG97EEB Wrote: You are not allowed to view links. Register or Login to view.Hi Mark,
I ran a combination GPT and Claude.. Hope it helps
Ed
Thanks for that.
When I look at these kinds of AI researches as with the previous stuff I tried with GPT and Gemini I don't know how to react. My initial reaction is that this is brilliant and very impressive. However when I start to work through the details I am not sure. First, I wonder to what extent it is feeding back to me in a slightly garbled/misinterpreted form what I have posted on Nick Pelling's blog as I have posted a lot about the ciphers that I have located. However, funnily enough I like it's confidence which I find encouraging. I find the reference to "Andrea Barbarigo’s private merchant cipher" intriguing. I have been in touch with Dr. Iordanou before, so I could follow that up. Paolo Bonavoglia says he has searched the Venice archives for diplomatic ciphers of that period thoroughly, so either he hasn't searched it as thoroughly as he thinks or there isn't any more to be found there. It seems to not have picked up on the 1397 Milanese cipher in the Chronicles of Lucca by Giovanni Sercambi as the earliest example of homophones we know of.
It is right that the State Archives of Mantua, Modena and Florence Archives need to be searched from top to bottom as they have already yielded much interesting materials, but there may be more to be found there.
It may be worth confirming that I have seen the following that it mentions:
1404–1418 enciphered correspondence
Jacobo de Faitinellis, Dino ser Paci, Iohannello Thieri, Guido da Pietrasanta, Nicolao da Moncicoli, Nicolao Arnolfini.
Foreign Correspondence, b. 25, sez. 3, Rome
Look for the 1426–1430 enciphered letters: Ricardo de Mutiliana, Pandolfo Malatesta, unsigned Rome letter, Francesco de Catalenis.
---
The tendency of these AI to hallucinate or just misunderstand what was previously been written makes me wonder whether it's references are genuine and correctly dated.I think it would be worth asking the AIs for the precise "shelfmark" for these ciphers and where they saw the references to them.
I like to think that I am not so proud that I care where a new lead comes from and that I will follow it up regardless.
Unfortunately the first rule of LLM is to always double and triple check, particularly before setting off on a trip to Italy
Edit: GPT confirmed these came from a Pelling blog post -
RE: Using AI to do research - Specifically online searching
eggyk > 4 hours ago
(9 hours ago)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.I am always looking for ways to find new examples of early 15th century ciphers. As those who are familiar with my online OneDrive archive know I have collected many images of examples of those and there are still some that I haven’t uploaded to my OneDrive. However, I continue to try to increase my collection. So, I have been experimenting with tools like Gemini and ChatGPT to see if they can help generate any new leads. I haven't found anything definitely interesting yet, although they have given me food for thought. However, I wonder if anyone else has experience doing this kind of research.
Something that I find useful is using AI to check supposed dead ends. For example, using AI to search for a name referenced in a manuscript. Sometimes AI will find other obscure parts of the internet where the same name was mentioned, including in situations were names have multiple variants (latin name, french name, german name for the same person).
A recent example I had was an irish nobleman who wrote a letter in french. AI found other variants of his name, which allowed me to re-search in digital libraries/google for any hits.
So, if you find a name referenced, or the name of a writing tradition, a location or specific church or abbey, or some other tidbit of info, AI can sometimes give you valuable leads where traditional searching comes up with nothing. -
RE: Using AI to do research - Specifically online searching
Stefan Wirtz_2 > 2 hours ago
Found AI (Google) severely hallucinating even with simple translations.
The first answer may be useful, but after that, „AI“ tries to guess the expectations of it‘s user.
This is not „Artificial Intelligence“, but more of „Artificial Empathy“.
AE intends to fulfill the Questioners dreams and is quite good with that, from the 2nd reply on, plus an ability to lie without any hesitation.
Sometime it helps to write that „AI“ is bullshitting in this moment, it even produces apologies.
But closing the tab or better, not using it all, saves some time.
I don‘t see much sense in those systems — all who are doing elaborated statistics and calculations: better have a look again. -
RE: Using AI to do research - Specifically online searching
JoJo_Jost > 42 minutes ago
gemini is a fake KI
