Category: CLE, Alternative Dispute Resolution (CLE), Bankruptcy and Debtor-Creditor (CLE), Business & Corporations (CLE), Commercial & Consumer Law (CLE), Constitutional Law (CLE), Criminal Law (CLE), Education Law (CLE), Elder Law (CLE), Employment and Labor Law (CLE), Environmental Law (CLE), Family Law (CLE), Government Law (CLE), Health Law (CLE), Intellectual Property Law (CLE), International Law (CLE), Law Practice Management (CLE), Litigation (CLE), Real Property Law (CLE), Taxation (CLE), Tort Law (CLE), Trusts, Wills & Estates (CLE) (show less)
|
This class is eligible for CLE Passport use. You will be able to select the CLE Passport as your payment method for no additional cost when registering for this course.
Credits: 2.0 General Credit hours*
*This class qualifies for the California Technology in the Practice of Law.
*This class qualifies for the NY Cybersecurity, Privacy and Data Protection - General.
*This class may also qualify for technology credit in other jurisdictions having such requirements. Please contact cle@dcbar.org with any additional questions.
Series Description: In an era where artificial intelligence (AI) is rapidly transforming the legal landscape, understanding its underlying mechanisms, strengths, and limitations is essential for attorneys. This new four-part series provides a comprehensive foundation for legal professionals who want to stay ahead of the AI curve and navigate its implications responsibly and effectively.
Class Description: This essential class in our Foundations in Modern Law: Specialized AI Practice Series offers attorneys a comprehensive understanding of language modeling – an area increasingly integral to legal work but often misunderstood. Beginning with a brief history, we’ll explore the progression from “bag-of-words” models to transformer-based architectures that power today’s advanced applications. Our faculty expert will unpack the major components of modern language models, including learned token embeddings, general pre-training, attention mechanisms, vision-language models, and the emergent functionalities that have significantly enhanced AI’s capabilities in language understanding and generation.
This class also addresses a critical knowledge gap: While attorneys are often expected to integrate language models into their workflows, few fully understand their inner workings or their strengths and limitations beyond a general concept of “predicting the next word.” We’ll demystify these mechanisms by examining the differences between recurrent neural networks (RNNs) and transformer neural networks – an understanding that is crucial for attorneys to appreciate the transformative impact of model architectures on data interpretation and output quality.
You will have access to guided, hands-on demonstrations using free, publicly available models, allowing you to experiment with key concepts at home. These exercises will illustrate the difference between explicit and implicit modeling capabilities of neural networks, providing a practical perspective on how language models process and respond to input data.
By the end of this class, you will have a well-rounded familiarity with each language model component, a deep understanding of the capabilities and limitations of RNNs and transformers, and insight into the specific ways these models can – and cannot – enhance legal tasks. This foundational knowledge is indispensable for attorneys who aim to use AI responsibly and strategically, with a clear view of its implications on legal practice. Those interested in this class may also want to attend How AI Really Works and What Models are Actually Modeling (On-Demand), Working Alongside AI Responsibly & Ethically (On-Demand) and Risks Baked into AI (On-Demand).
Faculty:
Todd Smith, D.C.’s Office of the Chief Technology Officer
Cosponsor:
D.C.’s Office of the Chief Technology Officer
Adding Registration, Please wait...
