VIENNA LEGAL HACKERS virtual Meetup 2_2023: “ChatGPT and Law”

The Vienna chapter of the global Legal Hackers community invites to the following hybrid meetup:

Title: “ChatGPT and Law”

In cooperation with ReMeP – Research Meets Practice and IRI§23 Rechtsinformatik Symposion.

Organization:
Dr. Stefan Eder, ReMeP Chair
Prof. Erich Schweighofer, IRI§ Program Management – University of Vienna

Format:
Welcome & Introduction
Keynote presentations
Closing panel discussion with Q&A session
(subject to change)

Topic 1: Large Linguistic Models (LLM) using the ChatGPT technology as an example. Initially, experts in information retrieval, text analysis and computational linguistics will explain the ChatGPT model; specifically how an answer is generated and why it often yields (at least in good part) impressive correct results.
What can ChatGPT do, how can it be used, what will it cost, is there already a German language model, what is the OpenAI research so much better than our own work on legal Natural Language Processing?
Keynote speech: (Informal) research group “Information Retrieval, Natural Language Processing, Entity Recognition and Similarity Search (Clustering) in Law”: Andi Rauber, Werner Winiwarter, Erich Schweighofer, Florina Piroi, Tobias Fink, Dieter Merkl, Michael Dietenbach and many others.

Lecturers: Florina Piroi, Tobias Fink (Team Prof. Rauber)

Large Language Models (LLM) have always been of interest to lawyers; the intensive effort in studying and in continuous learning shows the need to get some technological support here as well. Information retrieval professionals have always been aware of the potential for problems of not knowing about the text corpus and its language (for related research see Schweighofer, Wissensrepresentation und Rechtsinformatik, 1999, Bin/Torvald, Legal Information Retrieval, 1987, Turtle, Text Retrieval in the Legal World, 1995 etc.; most recently: see the work of Bernhard Waltl, LLI). On ChatGPT and language models: Frank Schilder/Jack Conrad, Exploring the Landscape of Large Language Models: ChatGPT and beyond, 2023. Jack Conrad is organizing a related workshop at ICAIL2023 in June 2023 in Braga PT.

Expert panel: Manz Verlag, LexisNexis, Verlag Austria, Linde Verlag, Cybly.tech Verlag, Weblaw, RIS/BRZ (invited)

Dr Wolfgang Pichler, Chief Evangelist, Manz Verlag
Dr Anton Geist LL.M, LexisNexis
etc. (is continuously updated)

Topic 2: The expertise of lawyers is to a large extent corpus-based, which changes dynamically every day (e.g. new legal provision or new decision). The (pre-trained) language model would have to access the corpus at the respective time of validity in order to get meaningful and relevant results. If this is not possible: what indicators are there for the lawyers that this result is risky?
If the language model is available in German: what specific legal questions arise (dynamic body of text, validity, applicability, metadata, etc.)?

Keynote speech: Jack Conrad, Consultant (früher: Thomson Reuters Labs)

Topic 3: Business Cases, Research Questions and Outlook
Finally, the obvious use case is discussed: the summary of a text, at least according to objective criteria, but also according to subjective goals (decisive for lawyers – what do I not know yet?). This already seems to work well; but the inclusion of metadata is still insufficient in any case. Evaluating accuracy and managing error is critical. Full accuracy is often not possible in law, but the risk must be manageable. Particular attention should be paid to the training of human-machine cooperation.

Moderation: Stefan Eder (Benn-Ibler Rechtsanwälte, cybly.tech)

In the meantime, stay up to date with our social media:

***https://twitter.com/VIELegalHackers ***

***https://www.linkedin.com/company/vie-legalhackers/?viewAsMember=true***

***https://www.facebook.com/viennalegalhackers***

The participation is free of charge. Registration will be required to receive link to streaming platform!!!

This event is made possible by our sponsors and partners: