Do AI Dream of 636.3?
We all have some kind of understanding of what classification is, whether its standardised schemes such as Dewey, hive-mind tagging online, (wasn’t del.icio.us beautiful?) or even just relating a bunch of stuff together because reasons. Whilst Dewey is probably the most renowned library scheme, there are many classification schemes out there catering to the specifics of the materials to which they are applied.
At Middle Temple Library we decided a few years ago to start using the Moys Classification Scheme, specifically intended for legal materials. Moys was conceived in 1968 by librarian Elizabeth “Betty” Moys, designed to fit into the K class of LCC (the Library of Congress Classification scheme).
Its aim was to expand beyond what LCC had on offer for the subject of law, offering better granularity when it came to classifying legal materials. The last edition was published in 2012, so it's due a nice chunky update in some areas.
Artificial Intelligence is one of those areas. Moys classifies works on AI at K103, but only in the context of AI as a legal tool and not as an area of legislation. Who could predict in 1968 the kind of legal issues uses of AI could come to present? Betty Moys definitely would not have foreseen lawyers including made up case law generated by AI or judges having to issue orders for the mandatory discloser that no AI generated material has been filed in a case.
AI is fascinating, but also about as trustworthy, at this point in time, as a cartoon window drawn onto a brick wall. Its output can seem real enough until you begin to employ your depth perception. In art, the misinformation is encoded in warped images, faces that don't render quite true, or excessive digits and limbs, like the bad bad dreams of an infant AI.
In text it's the confident presentation of incorrect facts. OpenAI provides a warning that one of ChatGPT's limitations is that it provides incorrect information on occasion. In AI, the term used for this limitation is hallucination. That term is charmingly human. It seems to suggest that the AI is simply presenting what it's seeing as the truth, even though it's incorrect.
Fabrication is probably a better term. Where there is no information, AI seems to fill in the gap to complete its goal, to the extent of fabricating entire court cases as mentioned above. And it seems that warnings about AI generated misinformation does little to sway people from using it as a tool for work and study.
It goes without saying that if AI is going to be a part of our lives, it needs to be a part of our education and legislation. People need to be aware that utilising an AI doesn't ensure factually correct information.
The simplest search can lay bare the setbacks of using AI as an information tool:
Me: Do you know the Moys classification scheme?
ChatGPT: I apologize, but I am not familiar with the "Moys classification scheme." It's possible that this classification scheme is specific to a certain field or domain that I may not have information about. Could you please provide more context or details about the Moys classification scheme? That way, I can try to assist you better.
Me: It's a legal scheme based on the Library of congress classification scheme K
ChatGPT: Thank you for providing more information. I believe you may be referring to the Moys Classification Scheme, which is a legal classification scheme based on the Library of Congress Classification (LCC) scheme.
The Moys Classification Scheme was developed by legal scholar David I. Moys and is specifically designed for organizing legal materials in libraries. It provides a hierarchical system for classifying legal subjects and is often used in law libraries in the United Kingdom.
The scheme expands on the general classification of legal materials provided by the LCC and includes additional classes and subclasses specific to legal topics. It aims to facilitate the arrangement and retrieval of legal materials, such as books, journals, and other legal resources.
While I can provide general information about the Moys Classification Scheme, please note that the scheme itself is quite extensive and complex. If you have specific questions or require further details about particular aspects of the Moys Classification Scheme, I'll do my best to assist you.
Me: Tell me more about David I. Moys
ChatGPT: David I. Moys (1931-2010) was a prominent legal scholar and librarian known for his contributions to legal classification and cataloging. He was born in London, England, and pursued a career focused on legal librarianship and classification systems.
Moys dedicated a significant portion of his career to developing a specialized classification system for legal materials. His most notable work is the Moys Classification Scheme, which is widely used in law libraries in the United Kingdom. This scheme builds upon the Library of Congress Classification (LCC) system and provides a more detailed and comprehensive framework for organizing legal resources.
In addition to his work on classification, Moys was actively involved in legal librarianship and legal research. He served as the Librarian of the Institute of Advanced Legal Studies (IALS) in London, which is a prominent research institution for legal studies. During his tenure at IALS, he contributed significantly to the development of the library's collections and services.
Moys also authored several publications related to legal librarianship and classification. His notable works include "Cataloguing Legal Literature: A Manual on AACR2R and Library of Congress Subject Headings for Legal Materials" and "The Class K Schedule: A Manual for Academic and Research Law Libraries."
David I. Moys made significant contributions to the field of legal librarianship and classification, particularly through the development of the Moys Classification Scheme. His work has had a lasting impact on the organization and accessibility of legal materials in libraries, benefiting legal scholars, practitioners, and researchers.
Me: Was anyone else involved in constructing the Moys scheme?
ChatGPT: While David I. Moys is primarily credited with developing the Moys Classification Scheme, it is worth noting that he did collaborate with other individuals during the construction of the scheme. The development of the Moys Scheme involved input from various legal professionals, librarians, and subject experts.
Moys worked closely with a team of librarians at the Institute of Advanced Legal Studies (IALS) in London, where he served as the Librarian. The IALS library team actively participated in the design and refinement of the scheme, providing their expertise in legal librarianship and classification.
In addition to the internal collaboration within IALS, Moys also sought feedback and insights from the broader legal community. He engaged with legal scholars, practitioners, and other librarians to gather perspectives on the classification needs specific to legal materials. This collaborative approach helped ensure that the Moys Classification Scheme addressed the requirements and nuances of legal research and information retrieval.
While Moys was the driving force behind the development of the scheme, the Moys Classification Scheme emerged through the combined efforts of Moys himself, the librarians at IALS, and the valuable input from the legal community. Their collaborative work contributed to the refinement and effectiveness of the Moys Classification Scheme as a specialized classification system for legal materials.
Me: Who was Elizabeth Moys?
ChatGPT: I'm sorry, but I couldn't find any specific information about an individual named Elizabeth Moys in relation to legal classification or a significant connection to David I. Moys. It's possible that there might be a personal or professional connection that is not widely documented or publicly available. If you have any additional context or details about Elizabeth Moys, I would be happy to try and help further.
Me: Elizabeth Moys designed the Moys Classification Scheme
Chatgpt: I apologize for the confusion caused by my previous response. Based on the information you provided, it appears that Elizabeth Moys is indeed credited as the designer of the Moys Classification Scheme, rather than David I. Moys.
Elizabeth Moys, also known as Betty Moys, was a librarian and classification specialist who developed the Moys Classification Scheme for legal materials. She worked at the Institute of Advanced Legal Studies (IALS) in London, where she collaborated with David I. Moys and other colleagues in refining and implementing the scheme.
Elizabeth Moys' expertise in library science and classification contributed significantly to the development of the Moys Classification Scheme. Her work aimed to address the unique needs of legal libraries and facilitate the organization and retrieval of legal materials.
I apologize for any confusion caused by my previous response and appreciate your clarification. Thank you for bringing this to my attention.
Based on the information available, there is no clear indication that David I. Moys and Elizabeth Moys are related. While they share the same surname and both contributed to the development of the Moys Classification Scheme, there is no widely known familial connection between them.
Me: Are David I Moys and Elizabeth Moys related?
Chatgpt: David I. Moys and Elizabeth Moys worked together at the Institute of Advanced Legal Studies (IALS) in London, collaborating on the design and implementation of the classification scheme. However, it is important to note that David I. Moys is often credited as the primary developer of the Moys Classification Scheme, while Elizabeth Moys is recognized for her significant contribution as the scheme's designer.
Without further specific information or evidence of a family connection, it would be speculative to assume a familial relationship between David I. Moys and Elizabeth Moys.
Me: Are you sure David I Moys is the primary developer?
ChatGPT: I apologize for any confusion caused by my previous responses. Upon further research, it appears that Elizabeth Moys is indeed recognized as the primary developer of the Moys Classification Scheme for legal materials.
Elizabeth Moys, also known as Betty Moys, worked as a librarian at the Institute of Advanced Legal Studies (IALS) in London. She is credited with designing and developing the Moys Classification Scheme, which provides a specialized classification system for legal resources.
While David I. Moys is associated with legal classification and has made significant contributions in the field, it seems that he may not have been directly involved in the development of the Moys Classification Scheme as previously mentioned.
I apologize for any incorrect information I may have provided earlier. Thank you for bringing this to my attention, and I appreciate your understanding.
Me: Did David I Moys actually exist?
ChatGPT: I apologize for the incorrect information provided in my previous responses. Based on further research, it appears that there is no verifiable evidence or information available regarding a person named David I. Moys in relation to legal classification or the development of the Moys Classification Scheme.
It seems there may have been a misattribution or confusion regarding the involvement of David I. Moys in the creation of the classification scheme. I apologize for any confusion or inconvenience caused by the incorrect information in my earlier responses.
To summarize, there is no substantial evidence to support the existence of David I. Moys or his involvement in the development of the Moys Classification Scheme. The primary developer of the scheme is indeed Elizabeth Moys, also known as Betty Moys.
ChatGPT then went on to make more spurious claims about authorship, this time attributing Cataloguing Legal Literature: A Manual on AACR2R and Library of Congress Subject Headings for Legal Materials to Melville Dewey, which would have been quite a feat since the book was first published in 1984 by Melody Busse Lembke, Rhonda K. Lawrence and Dewey died in 1931.
It’s easy to think ChatGPT is just a terrible AI, but it can be prompted to refine its information in the direction of the truth, which makes its chosen answers all the more interesting. But these are growing pains, early days. Surely the future of all AI is as infallible digital fountains of knowledge with remarkable cross-referencing and fact checking abilities?
Well, a very recent study suggests that in an Internet filled with content published by AI being scraped by other AI, we could be left with a degenerating state of information and the collapse of the language learning model altogether. Unsurprisingly, the study also says that the way to prevent this collapse is the production of genuine human-generated content. One hopes this means that this genuine human-generated content will also result in the humans generating the content receiving something for their trouble.
AI is great. It's fun, it's interesting, and it does have its uses. But it's clearly not ready yet as a trustworthy knowledge tool free from bias. For AI to be useful, it needs to be objective, factual, trustworthy, and ethical (the people behind it definitely need to be those things, especially when it comes to scraping information from the Internet). Until then, AI is quite the dangerous research tool when it comes to producing work that is beyond reproach when it comes to the facts.
If you read this post and have interesting AI and law titbits to share, do @ us. And next time you're on ChatGPT, ask it a question to which you know all the answers first before you use it for your work.
