GENERATIVE

ARTIFICIAL INTELLIGENCE

A CRITICAL ENGAGEMENT TOOLKIT

ARTIFICIAL INTELLIGENCE IN k-12

As artificial intelligence (AI) becomes more ubiquitous in the world around us, schools can be a crucial site where students can learn how AI works, gain the skills to effectively use AI-based technologies (or choose not to), design AI systems, and critically examine AI and its effects. Since 2022, the public release of ChatGPT and other similar platforms have made it widely possible for students and teachers to directly interface with AI tools in education. AI is also being embedded into features of existing platforms used in schools. Questions about how AI might support or detract from teaching and learning, change the nature of assessment, and impact the future of work are being asked urgently.

GENERATIVE ARTIFICIAL INTELLIGENCE

There are many different types of Artificial Intelligence. This toolkit focuses on generative AI (genAI), which can be used to generate text, images, code, video, audio and other forms of content.

Generative AI is not intelligent. However, it is able to learn from patterns of data it was trained on to create content that can seem remarkably similar to human understanding, creation and communication. This capacity makes it a powerful tool that is having a ripple effect on society and on education.

​

​

A CRITICAL ENGAGEMENT TOOLKIT

Technology is not neutral and neither are the societies in which we live. Critical engagement involves being thoughtful about our approach to technology, deliberately reflecting on our own assumptions and actions, questioning taken-for-granted ideas and grappling with technology’s effects.

This critical engagement toolkit is meant to go beyond the ‘how-to’ of using genAI tools. It provides a scaffolded way to build the habits of mind for engaging in dynamic processes of reflection and action as they relate to genAI.

​

​

​

​

​

HOW THE TOOLKIT WORKS

This toolkit is designed to support metacognitive thinking about how and why different aspects of genAI matter for specific uses or contexts.

The toolkit is organized around two key dimensions: EXPLORE THE TECH & CHECK IN WITH YOURSELF. Each dimension offers four prompts and some central guiding QUESTIONS that can be used to generate conversation and reflection.

The remainder of the toolkit includes SUPPORTING SLIDES that offer additional information and guidance for deeper reflection, and can be adapted for use in any order.

​

​

​

​

​

​

​

​

WAYS TO USE THIS TOOLKIT

genAI:

CRITICAL ENGAGEMENT

TOOLKIT

GENERATIVE AI: A CRITICAL ENGAGEMENT TOOLKIT

GENERATIVE AI: A CRITICAL ENGAGEMENT TOOLKIT

Consider limitations

What are genAI’s limitations

and why do they matter?

Use AI ethically

Is my use of genAI ethical?

EXPLORE THE TECH

CHECK IN WITH YOURSELF

EXPLORE THE TECH

EXPLORE THE TECH

​

GenAI is distinct from other technologies and especially powerful, and can even feel like “magic” including because:

GenAI can appear to have understanding and intelligence

GenAI can appear social and relational

We tend to attribute human characteristics to GenAI

GenAI can influence our thoughts, feelings, judgements and actions

Understanding the tech helps to remove some of the “magic” and allows us to use these tools more effectively and critically, mitigate some potential harms, and advocate for the responsible development of these tools.

​

​

​

​

​

​

​

​

​

​

UNDERSTAND THE TECH

How does genAI technology work?

EXPLORE THE TECH

• GenAI is effective because it uses Neural network technology which is inspired by complex layers of synaptic connections in the human brain.

• GenAI works by using statistical techniques to predict the next most likely word in a sentence.

The complexity of these models makes them black boxes, which means that no one can explain exactly how the model came to a specific response.

GenAI relies on training an AI model on vast amounts of data, and on extensive human labor.

While AI is very good at mimicking human-created content, it does not actually have real intentions or human-like understanding.

AI can also present false information as though it were fact. These “hallucinations” can be especially tricky to identify because the information is plausible and presented very confidently.

LOOK INTO THE SOURCE

​

​

What data, companies, and natural resources are behind the scenes?

​

​

EXPLORE THE TECH

INTERROGATE DATA PRIVACY

GenAI platforms platforms collect and monetize enormous amounts of data (like our browsing history, the apps we use, and our location), often without our knowledge or consent. Being aware of what data is being collected and how it might be used is a first step in have some degree of agency over our data.

​

Two important guidelines for educators include:

​

​

​

What kind of data is this tool collecting and what happens to it?

Remind students that their prompts and interactions within a genAI platform are NOT private. Using genAI can often feel like talking with a friend or a real person, which may lead us to reveal more information than we otherwise would. In some cases, educators can view a student’s chat history, company staff can actively review user interactions within a platform, or data can be shared with third parties, law enforcement, and even subject to data breaches.

Be wary of entering ANY student data or profile information into tools that don’t comply with necessary data privacy agreements. Even when we anonymize data by removing personally identifiable information, the magnitude of data that is collected about individuals across platforms and applications can make them easily vulnerable to re-identification, targeted advertising and other data harms.

EXPLORE THE TECH

CONSIDER THE LIMITATIONS

What are genAI’s limitations and why do they matter?

GenAI often confidently asserts incorrect information, known as “hallucinations.” When might this be harmless? When might it have serious consequences? How could I actively anticipate and address this limitation?

EXPLORE THE TECH

• AI reflects the biases of the people who designed and trained it, and biases that are part of the dataset it was trained on. Whose voices, experiences and perspectives are most likely to be represented or missing? How might this affect its output?

EXPLORE THE TECH

• GenAI tools are prone to making mistakes (or being misused) in ways that have real-life consequences, with few mechanisms for accountability. When something goes wrong, what can I do? Which people and networks can I lean on for help and support?

LEARN ABOUT HOW AI IS CHANGING

AI technologies and products are changing every day. What is true about AI tools today may no longer be true tomorrow. The pace at which technology changes can be overwhelming. �

What new tools or updates to existing tools do I need to know about to inform my use of genAI?�

What resources, people, and networks will I lean on to continue to learn about AI?

​

How do my answers change when I revisit the questions in this section over time?

​

​

​

​

​

​

​

​

​

​

​

​

​

​

KEEP EXPLORING THE TECH

CHECK IN WITH YOURSELF

CHECK IN WITH YOURSELF

​

When we interact with and use genAI technologies, we implicitly bring our own feelings, experiences, values, and perspectives to it. While genAI platforms or tools influence and shape our use, we also have agency in making decisions before, during and after engaging with generative AI.

​

Part of using genAI tools effectively and making sound judgements about them includes explicitly surfacing:

​

What we bring to genAI tools

Whether and when to use genAI tools

How to use genAI tools

​

​

​

​

​

NOTICE YOUR FEELINGS

Whether or not we are using these tools, AI can bring up a whole host of feelings. Feelings are an important part of human intelligence and wisdom, so noticing and exploring those feelings can indicate something important to us about our own experiences, perceptions, and needs around AI.

​

These questions may be helpful for disentangling our feelings:

​

​

​

What feelings does genAI bring up for me?

What feelings does the idea of AI bring up for me? Why? Has that changed for me over time?

Am I good enough to do this without AI?

1

2

3

What do I need to do? How am I feeling about this task? How would the use of genAI change my feelings about this task?

How do I feel about sharing or using something that AI has helped me create?

CHECK IN WITH YOURSELF

How will I feel when I describe my use of AI to friends, supervisors, teachers, etc?

ASSESS THE TOOL

CHECK IN WITH YOURSELF

Is genAI right for my needs?

• A range of factors such as access (including cost, age limits, etc.), complexity, ease of use, and our own level of skill/experience can determine whether we use a specific tool. Is this tool worth investing my time (work time, instructional time) to learn and use, based on the potential benefits?

​

• There are differences in capability between different AI tools and between their free and paid versions (e.g., cutoff date for training data, file upload ability). GenAI tools are also constantly changing as new versions of the platforms or the underlying language model become available. What are the specific features and affordances of the genAI tool that I am using?

• GenAI tools work well because they can recognize and mimic patterns, so they can be very good at things like leveling or summarizing texts for example, but not very good at things like providing factual information or making judgements that rely on professional experience. What types of tasks would I use (or not use) specific genAI tools for? Why or why not?
• ​

ENGAGE ACTIVELY

How we use genAI tools can either support our growth, learning, and agency, or diminish it. However, the difference between these ways of using tech tools is not always easy to discern.

​

That is why it is helpful to continuously and explicitly ask ourselves about what we are using genAI for, which skills we are building, which tasks we are choosing to offload, and why:

shifting time?

Am I using genAI in a way that supports my learning and agency?

reflecting?

using my creativity?

making decisions?

Am I actively...

thinking critically?

developing a skill?

finding new ideas and inspiration?

problem solving?

connecting with others?

CHECK IN WITH YOURSELF

saving time?

USE AI ETHICALLY

CHECK IN WITH YOURSELF

• Checking on school/district policy (eg. acceptable use) can help us remain accountable to the institutions we are part of.
• ​

We must model transparency and reflection around purposeful genAI use (eg. including a statement about how AI was used).

In deciding whether or how to use genAI, it is valuable to gauge our internal compass. Our gut instincts can provide valuable information.

It is important to be aware that there is an ongoing debate about genAI, copyright and intellectual property.

Is my use of genAI ethical?

If we choose to use genAI in any part of our work, we are still fully responsible for the end product and its effects.

• AI detectors are often inaccurate, unreliable and biased. They may also use student work for training and advertising purposes.

​

ALLOW YOUR APPROACH TO EVOLVE

Every time we use genAI — or consider using genAI — we have an opportunity to learn more deeply about this impactful technology and about our individual and collective relationships to it.

​

How are my feelings, thoughts, needs, and skills evolving?

​

Are my conversations about AI expanding or reinforcing my perspectives?

​

How do my answers change when I revisit the questions in this section over time?

​

​

​

​

​

​

​

​

​

​

​

​

KEEP CHECKING IN WITH YOURSELF

OUR TEAM

This toolkit was created by a team of edtech specialists and education researchers at the Collaborative for Educational Services, a non-profit in Western Massachusetts dedicated to supporting teaching and learning in our region and beyond.

Although we each bring deep expertise in our own fields, we are not technical experts on AI. As people committed to public education and equity in our schools, we seek to model that it is important for all of us to be actively and collaboratively engaged in these critical conversations.

Design by Lauren Parent.

Casey Daigle

Dr. Rebecca Mazur

Suzanne Judson-Whitehouse

OUR PROCESS

The iterative process of creating this toolkit involved extensive research and intensive dialogue and collaboration with each other, as well as valuable feedback from educators.

A long list of people and organizations have also shaped our thinking about AI, including:

Emily Bender, Ruha Benjamin, Meredith Broussard, Kate Crawford, Eric Curts, Timnit Gebru, Cathy O'Neil, Arvind Narayanan, Safiya Umoja Noble, John Spencer, Torrey Trust, AI Breakfast, Allied Media Projects, Center for Humane Technology, Civics of Technology, DAIR, ISTE, and MIT Day of AI, among others.

USE OF AI STATEMENT

Generative AI was not used at any stage to generate or edit/refine the content (text, icons or design) in this toolkit.

However, we used generative AI tools extensively in order to learn more about them, explore their capabilities, reflect on our own feelings, discuss our process of using these tools, as well as to deepen our own understanding of various ideas and concepts related to artificial intelligence. The tools that were explored include:

​

​

• Perplexity
• Perplexity Pro
• Bard/Gemini
• Magicschool
• Midjourney

• ChatGPT-3.5
• ChatGPT-4
• Claude
• Dall-E
• Diffit

• Embedded tools:
• Adobe
• Canva
• Padlet

​

How to cite this toolkit:

Generative Artificial Intelligence Slide Deck by Collaborative for Educational Services is licensed under CC BY-NC-ND.

​

As educators, we passionately believe that copyright should both protect intellectual property and inspire new creations. This slide deck is licensed under CC BY-NC-ND which includes the following elements:

BY: Credit must be given to the creator.

NC: Only noncommercial uses of the work are permitted.

ND: No derivatives or adaptations of the work are permitted.

Other Uses:

If you would like to use this Toolkit in a way not covered by the Creative Commons license outlined above, please contact us to discuss your proposed use. We are open to other uses that align with our intentions, research, and mission and welcome a conversation.