1 of 36

Digital Equity Webinar Series

Using AI to Foster Critical Thinking about Bias, Ethics, and Data

2 of 36

Welcome and Introductions

3 of 36

Guest Speaker - Caleb Wilson

Caleb Wilson is the District Coach of Instructional Technology and Innovation for Worcester County Public Schools in Maryland, where he leads efforts to integrate AI into classrooms with a strong emphasis on ethics, bias, and data integrity. As the head of the AI4WCPS initiative, Caleb works with educators to foster critical thinking around AI's societal impacts and ethical use, ensuring responsible and inclusive technology practices. He also collaborates with the Maryland Center for Computing Education, guiding teachers statewide in the ethical integration of AI in education.

4 of 36

Panelist - Charlotte Avery

Charlotte Avery is the Program Impact Manager for the Institute for Trustworthy AI in Law and Society (TRAILS) and the Artificial Intelligence Interdisciplinary Institute at Maryland (AIM), both at the University of Maryland. In this role, she leads the design, execution, and management of programs centered on participatory design, with a focus on educational and workforce development initiatives that promote trustworthy and responsible AI. With over 15 years of experience in K-12 and higher education, Charlotte is a seasoned expert in curriculum design, teacher preparation, coaching, event management, and strategic program development. Charlotte is currently pursuing a PhD in Information Studies at the University of Maryland, where her research examines the effects of emerging technologies on education systems, with a special focus on AI’s influence on access, equity, and the experiences of historically underrepresented populations in schools. In her free time, Charlotte enjoys building LEGO models, staying active, practicing self-care, and spending quality time with her family and friends.

5 of 36

Panelist - Ron Nicholson

Ron Nicholson is a Co-Founder and currently serves as Executive Director for BCAUSEICAN INC. BCAUSEICAN is a 501c3 nonprofit located in Greenbelt, MD that provides STEM education to students and adults in underserved communities. BCAUSEICAN conducts computer coding classes for grades K-12 and Artificial Intelligence classes to grades 6-12. BCAUSEICAN also conducts Adult Technology classes in English and Spanish. BCAUSEICAN is a member of the recently formed Prince George’s County Artificial Intelligence Task Force. Mr. Nicholson has spent most of his 30 year computer engineering career working for Georgetown University in the Main Campus Information Technology department. His Operational Team was responsible for bringing virtualization technology to Georgetown University in 2008. Mr. Nicholson studied Computer Science at N.C. State University.

6 of 36

Using AI to Foster Critical Thinking about Bias, Ethics, and Data

Speaker Presentation

7 of 36

Using AI to Foster Critical Thinking about Bias, Ethics, and Data

Friday, December 6th, 2024

Caleb J. Wilson

8 of 36

Caleb Wilson

  • District Coach of Instructional Technology and Innovation at Worcester County Public Schools
  • Masters in Curriculum and Instruction, concentrating in Educational Technology from Salisbury University.
  • 8 years of Chemistry and science education for 7-12
  • Founder of AI4WCPS - Artificial Intelligence Steering Committee for AI Integration into WCPS
  • AI Presenter for Maryland Center for Computing Education
  • Semi-Finalist for AI Educator Award hosted by MagicSchool.ai

9 of 36

Brief Agenda

01

02

04

03

05

06

AI Refresh

Data & Bias

Good Pedagogy

Equity & Access

Deepfakes & AI Literacy

Vision for the Future

10 of 36

AI Refresh

01

11 of 36

What is Artificial Intelligence?

Artificial intelligence is a field of science concerned with building computers and machines that can reason, learn, and act in such a way that would normally require human intelligence or that involves data whose scale exceeds what humans can analyze. �

  • seeing and hearing things
  • Understanding and responding to language
  • making decisions
  • translating languages
  • coding
  • recognizing patterns
  • and more . . .

12 of 36

What is Artificial Intelligence?

Artificial intelligence is a field of science concerned with building computers and machines that can reason, learn, and act in such a way that would normally require human intelligence or that involves data whose scale exceeds what humans can analyze. �

  • seeing and hearing things
  • Understanding and responding to language
  • making decisions
  • translating languages
  • coding
  • recognizing patterns
  • and more . . .

Creating machines that can do human things.

13 of 36

How does Artificial Intelligence work?

  • Explanation of AI systems:

AI systems use special computer programs called algorithms to complete tasks.�

  • Use of data:

These algorithms learn from lots of information called data to find patterns and make predictions.�

  • More data, better performance:

The more data an AI system has, the better it can do its job.

14 of 36

How does Artificial Intelligence work?

  • Explanation of AI systems:

AI systems use special computer programs called algorithms to complete tasks.�

  • Use of data:

These algorithms learn from lots of information called data to find patterns and make predictions.�

  • More data, better performance:

The more data an AI system has, the better it can do its job.

Machines take data, apply algorithms, make decisions.

15 of 36

What is generative AI? �What’s a Large Language Model?

A generative AI model can take what it has learned from the examples it’s been shown and create something entirely new based on that information.

Hence the word “generative!” Large language models (LLMs) are one type of generative AI since they generate novel combinations of text in the form of natural-sounding language.

There are also generative AI tools to generate music, pictures, videos, powerpoints, spreadsheets, code, and even interactive virtual worlds.

16 of 36

What data is ChatGPT trained on?

GPT-4 is trained on a massive dataset of text data from the internet, which includes web pages, books, articles, and other sources of written language up to October 2023*. The model uses this data to learn the patterns and relationships between words, phrases, and sentences, so that it can generate human-like responses to text inputs and is further refined using reinforcement learning with human feedback (RLHF) that improves quality and safety.

17 of 36

Data & Bias

02

18 of 36

19 of 36

Data Bias

20 of 36

Equity & Access

03

21 of 36

22 of 36

AI Model

Developer

Cost

Signup Process

Capabilities

Copilot�

Microsoft

Free, Only 30 chats per session and 300 per day total.

Copilot Pro, $30/per month.

Integrated with Microsoft's search engine; Google Chrome and Safari

Can search the web and incorporate the results into its answers. It can also generate images with DALL·E 3 Runs GPT4 Turbo

ChatGPT

OpenAI

Free for 3.5, Plus version for $20/month��Unlimited conversations in GPT-3.5-turbo, 50 per 3hrs with GPT-4

Accessible through OpenAI’s website and available as a stand alone app (iOS and Android)

Can search the web, access to plug-ins for expanded features, Can generate images with DALL-E 3.Can accept files and photos to analyze.

*Claude

Anthropic

Free, Paid version for $20/month

*Now Claude 3

Accessible through a chat interface via Anthropic’s website

Proficient in advanced reasoning, vision analysis, code generation, and multilingual processing. It can tackle complex cognitive tasks, analyze static images, create and debug code, and handle real-time language translation and content creation.

Gemini

Google

Free, Paid version $20/month. Gemini Pro/Ultra

Accessible through Google’s platform

Can pull information from the internet to provide text-based answers to user questions. It can also include images in its answers. Uses Gemini Pro

Pi.ai

Inflection AI

Free

Can make an account or try for free with no chat history

Can generate original and coherent responses on the fly, based on the context and the user’s input. It can also express emotions and empathy, using natural language and emojis.

23 of 36

AI Model

Developer

Cost

Signup Process

Capabilities

Copilot�

Microsoft

Free, Only 30 chats per session and 300 per day total.

Copilot Pro, $30/per month.

Integrated with Microsoft's search engine; Google Chrome and Safari

Can search the web and incorporate the results into its answers. It can also generate images with DALL·E 3 Runs GPT4 Turbo

ChatGPT

OpenAI

Free for 3.5, Plus version for $20/month��Unlimited conversations in GPT-3.5-turbo, 50 per 3hrs with GPT-4

Accessible through OpenAI’s website and available as a stand alone app (iOS and Android)

Can search the web, access to plug-ins for expanded features, Can generate images with DALL-E 3.Can accept files and photos to analyze.

*Claude

Anthropic

Free, Paid version for $20/month

*Now Claude 3

Accessible through a chat interface via Anthropic’s website

Proficient in advanced reasoning, vision analysis, code generation, and multilingual processing. It can tackle complex cognitive tasks, analyze static images, create and debug code, and handle real-time language translation and content creation.

Gemini

Google

Free, Paid version $20/month. Gemini Pro/Ultra

Accessible through Google’s platform

Can pull information from the internet to provide text-based answers to user questions. It can also include images in its answers. Uses Gemini Pro

Pi.ai

Inflection AI

Free

Can make an account or try for free with no chat history

Can generate original and coherent responses on the fly, based on the context and the user’s input. It can also express emotions and empathy, using natural language and emojis.

OUTDATED

Just go to their websites!

24 of 36

Does everyone have access?

  • Access to Tools
  • Knowledge to use tools
  • Ethics and AI literacy
  • $ for best quality.
  • Accessibility
  • Parental Knowledge
  • Teacher Knowledge
  • Classroom Experience

25 of 36

Good Pedagogy

04

26 of 36

27 of 36

What is good pedagogy?

How do you know when you taught a good lesson?

How do you know when a lesson is good for your students?

What is the teacher’s role?� - grading� -feedback� -decision making

AI Teaching/Learning Loop

28 of 36

Deepfakes & AI Literacy

05

29 of 36

AI Image Generation

30 of 36

What is real?

How do we verify information?

-deepfakes

-“plagiarism”

AI Literacy

31 of 36

Vision for the Future

06

32 of 36

How do we address all these challenges?

Align our uses of AI to our vision for the future.

Establish our values and use AI in ways that support and uphold them.

AI in our world today requires us to think about the future, both in the context of our students’ lives and in broader society.

33 of 36

THANK YOU!

CREDITS: This presentation template was created by Slidesgo, and includes icons by Flaticon and infographics & images by Freepik

34 of 36

Using AI to Foster Critical Thinking about Bias, Ethics, and Data

Panel Discussion

35 of 36

Questions

36 of 36

Thank You