1 of 26

01 - Summer Talks

General overview of generative AI

We’ll get started soon!

2 of 26

Setting the stage

3 of 26

What is AI?

4 of 26

What is AI?

Terminology for today

LLMs
Frontier/Foundational Models
Context Window
Tokens
RAG

5 of 26

November 30, 2022

GPT 3.5

The AI Big Bang

May 13, 2024

GPT 4o

March 4, 2023

GPT 4*

6 of 26

$1.3 Trillion

7 of 26

Why?

Why am I spending so much time on AI literacy and proficiency?

For one, we are in a unique situation here at the Law School. There is an education component. How does it fit pedagogically into your courses, how can you have meaningful conversations with your students about the learning process. Where does AI fit in? Where should it be avoided? What is it good at? Where will it get you in trouble.

We also have a duty, in my opinion, to teach our students how to use and best think about these tools. I was just forwarded recently an email from Bloomberg Law that Law Firms are offering training programs for summer associates on these tools. This will be a skill that employers look for from future employees starting right now. AI is constantly changing and is in the news a lot. This morning’s brief from the Financial Times discussed issues around antitrust watch dogs growing concern over monopoly issues with the top AI companies.

Research tools like Lexis and Westlaw are creating specially trained AI models and “agents” to synthesize data with natural language, other specialized tools like Harvey AI and others are being sold to large firms. Foundational models from the big tech companies Microsoft, Amazon, Google, OpenAI, Anthropic, etc. have sandboxed or local environments that can be packaged and trained on your own private data for better results.

These future lawyers no matter their area of expertise will be better suited to understand how these work at a high level, and those that go into areas where they may be regulating this industry should have a much deeper understanding of what it is they are regulating.

The other reason is that knowledge about these tools is key. It reduces or eliminates fear, it helps you be more productive with these tools as you explore their capabilities and limitations. It allows you to see it for what it is. A co-intelligence or an assistive intelligence. The path forward isn’t us vs. AI. We will move forward as a society and advance socially, culturally, economically, and more because we partner with AI.

Humans, I believe will always need to remain in the loop. In charge of creativity and be the catalyst for asking the right questions to get answers to the solutions we’re looking for.

8 of 26

Why?

Why am I spending so much time on AI literacy and proficiency?

For one, we are in a unique situation here at the Law School. There is an education component. How does it fit pedagogically into your courses, how can you have meaningful conversations with your students about the learning process. Where does AI fit in? Where should it be avoided? What is it good at? Where will it get you in trouble.

We also have a duty, in my opinion, to teach our students how to use and best think about these tools. I was just forwarded recently an email from Bloomberg Law that Law Firms are offering training programs for summer associates on these tools. This will be a skill that employers look for from future employees starting right now. AI is constantly changing and is in the news a lot. This morning’s brief from the Financial Times discussed issues around antitrust watch dogs growing concern over monopoly issues with the top AI companies.

Research tools like Lexis and Westlaw are creating specially trained AI models and “agents” to synthesize data with natural language, other specialized tools like Harvey AI and others are being sold to large firms. Foundational models from the big tech companies Microsoft, Amazon, Google, OpenAI, Anthropic, etc. have sandboxed or local environments that can be packaged and trained on your own private data for better results.

These future lawyers no matter their area of expertise will be better suited to understand how these work at a high level, and those that go into areas where they may be regulating this industry should have a much deeper understanding of what it is they are regulating.

The other reason is that knowledge about these tools is key. It reduces or eliminates fear, it helps you be more productive with these tools as you explore their capabilities and limitations. It allows you to see it for what it is. A co-intelligence or an assistive intelligence. The path forward isn’t us vs. AI. We will move forward as a society and advance socially, culturally, economically, and more because we partner with AI.

Humans, I believe will always need to remain in the loop. In charge of creativity and be the catalyst for asking the right questions to get answers to the solutions we’re looking for.

9 of 26

Why?

Why am I spending so much time on AI literacy and proficiency?

Second, we are in a unique situation here at the Law School. There is an education component. How does it fit pedagogically into your courses, how can you have meaningful conversations with your students about the learning process. Where does AI fit in? Where should it be avoided? What is it good at? Where will it get you in trouble.

We also have a duty, in my opinion, to teach our students how to use and best think about these tools. I was just forwarded recently an email from Bloomberg Law that Law Firms are offering training programs for summer associates on these tools. This will be a skill that employers look for from future employees starting right now. AI is constantly changing and is in the news a lot. This morning’s brief from the Financial Times discussed issues around antitrust watch dogs growing concern over monopoly issues with the top AI companies.

Research tools like Lexis and Westlaw are creating specially trained AI models and “agents” to synthesize data with natural language, other specialized tools like Harvey AI and others are being sold to large firms. Foundational models from the big tech companies Microsoft, Amazon, Google, OpenAI, Anthropic, etc. have sandboxed or local environments that can be packaged and trained on your own private data for better results.

These future lawyers no matter their area of expertise will be better suited to understand how these work at a high level, and those that go into areas where they may be regulating this industry should have a much deeper understanding of what it is they are regulating.

Another reason is that knowledge about these tools is key. It reduces or eliminates fear, it helps you be more productive with these tools as you explore their capabilities and limitations. It allows you to see it for what it is. A co-intelligence or an assistive intelligence. The path forward isn’t us vs. AI. We will move forward as a society and advance socially, culturally, economically, and more because we partner with AI.

Humans, I believe will always need to remain in the loop. In charge of creativity and be the catalyst for asking the right questions to get answers to the solutions we’re looking for.

10 of 26

Why?

Why am I spending so much time on AI literacy and proficiency?

Second, we are in a unique situation here at the Law School. There is an education component. How does it fit pedagogically into your courses, how can you have meaningful conversations with your students about the learning process. Where does AI fit in? Where should it be avoided? What is it good at? Where will it get you in trouble.

We also have a duty, in my opinion, to teach our students how to use and best think about these tools. I was just forwarded recently an email from Bloomberg Law that Law Firms are offering training programs for summer associates on these tools. This will be a skill that employers look for from future employees starting right now. AI is constantly changing and is in the news a lot. This morning’s brief from the Financial Times discussed issues around antitrust watch dogs growing concern over monopoly issues with the top AI companies.

Research tools like Lexis and Westlaw are creating specially trained AI models and “agents” to synthesize data with natural language, other specialized tools like Harvey AI and others are being sold to large firms. Foundational models from the big tech companies Microsoft, Amazon, Google, OpenAI, Anthropic, etc. have sandboxed or local environments that can be packaged and trained on your own private data for better results.

These future lawyers no matter their area of expertise will be better suited to understand how these work at a high level, and those that go into areas where they may be regulating this industry should have a much deeper understanding of what it is they are regulating.

Another reason is that knowledge about these tools is key. It reduces or eliminates fear, it helps you be more productive with these tools as you explore their capabilities and limitations. It allows you to see it for what it is. A co-intelligence or an assistive intelligence. The path forward isn’t us vs. AI. We will move forward as a society and advance socially, culturally, economically, and more because we partner with AI.

Humans, I believe will always need to remain in the loop. In charge of creativity and be the catalyst for asking the right questions to get answers to the solutions we’re looking for.

11 of 26

Jagged Frontier

12 of 26

13 of 26

14 of 26

What tools are currently available?

LLM Chat

ChatGPT
Claude
Gemini
Copilot
Meta AI

Image Generation

Midjourney
Dalle (OpenAI)
Imagen (Google)
Firefly (Adobe)
Stable Diffusion
Magnific

Audio Creation

ElevenLabs
Suno
Udio
SoundFx

15 of 26

What tools are currently available?

Video Generation

Sora
Pika Labs
Veo
Runway

Industry Specific tools

and “wrappers”

Lexis+ AI
Harvey AI
Westlaw
Khan Academy
Google’s NotebookLM
Perplexity
Too many others to name

16 of 26

How am I using AI?

Quickly iterating on ideas
Translation
Speech to text
Learning topics
FUN!

Summarizing content
Expanding ideas
Scaffolding content
Quick answers to questions
FUN!

General Purpose AI - LLM Chat Tools

Copy Editing
Content Creation
Talking through problems
Coding
Automating tasks
Reframing ideas or topics
FUN!

17 of 26

How am I using AI?

Assets for presentations
Assets for videos
Fun!

Audio Generation

Riffing on musical ideas
Creating music or sound effects for tutorial videos
Translation
FUN!

Image Creation

18 of 26

Thoughts From Teaching in Excellence Day

There is still a lot of confusion around generative AI
We need continued conversation to educate our community
Students Crave guidance
Students are using these tools
Faculty, staff and students that learn these tools and figure out useful applications will start producing better work and save time in doing so

19 of 26

GenAI is...

Weird

20 of 26

GenAI can be...

Scary

21 of 26

GenAI is not without its...

Faults

22 of 26

GenAI is already...

Useful

23 of 26

Don’t forget GenAI can be...

FUN!

24 of 26

What’s Next?

AI is just getting started. GPT-4o is amazingly helpful. Not without its flaws. Even if we only get a few more percentage points better, that’s still pretty good imo.
Microsoft says we’re still on the lower end of a hockey stick
Sam Altman says GPT-4 sucks
Estimations say up to 90% of content on the internet will be generated fully or assisted with AI technologies
Data, Data, Data, if you haven’t already, you’ll probably be asked to have scholarly work included in AI training sets in the near future
New training modalities will include vision, audio, and other senses to help with a better world view.
Specialized models will surpass human experts in their field.
Human + AI will unlock new and exciting things

25 of 26

What’s Next?

What’s next for this series?

More AI literacy - each workshop will illuminate how they work and key things to know
Prompt workshop
How to teach with AI
Assignments in the age of AI
Creating custom chat assistants

26 of 26

Demo Time!