SST Critical Awareness of Generative AI

What is Generative AI?

Generative Artificial Intelligence (GenAI) is just one type of model that exists under the umbrella of artificial intelligence. These large language models (LLMs) are trained on vast amounts of data to understand and process human language, and as a result, can produce text, audio, video, computer code, and other types of information.

To help us better understand how GenAI functions, let’s look at one model, ChatGPT Edu, which is available to all students, faculty, and staff at ASU.

I can give ChatGPT a prompt that asks it to create something.
ChatGPT runs this prompt through all the information that has been used to train it, which in the case of GenAI, is often the work of other human beings
ChatGPT creates something that it has determined meets my expectations.

This is a simplified explanation for a very complex process, but it highlights one of the most important aspects of GenAI that users should understand: the role that humans play in this process. The humans who created ChatGPT trained it using the work of other human beings so that a different human could ask it to create work that looks like what the other human beings created.

Understanding the impact LLMs have on humans is key to developing a critical awareness of generative AI.

Generative AI and the Democratization of Learning

What is the democratization of learning?

Democratization of learning refers to the ways in which education and knowledge are made accessible to all learners to the greatest extent possible. Within the context of generative AI, ASU recognizes the capability of AI models to make learning and engagement more inclusive for all members of its diverse population.

How does generative AI make learning more accessible and inclusive?

Generative AI models can facilitate understanding of new or complex learning materials.

GenAI has the ability to:

Summarize and analyze written text to help students better understand course materials.
Create practice questions and quizzes using uploaded materials to check comprehension.
Transcribe meeting notes to make locating important information easier.

Generative AI models can help students prioritize tasks and manage heavy workloads

GenAI has the ability to:

Make a study schedule based on due dates, learning materials and other priorities like student clubs, rehearsals, athletic practices and games, and work schedules
Break down assignments into steps based on assignment details and rubrics

Generative AI models help students who may need additional support

GenAI has the ability to:

Outline complex content to minimize cognitive load
Translate content into a student’s first language to help with comprehension of new knowledge
Create questions and talking points to help students prepare for class discussions or meetings with an instructor

Generative AI and a Critical Consciousness of AI Systems

What is a critical consciousness of AI Systems?

Critical Consciousness (CC) is a pedagogical approach introduced by Paulo Freire. Simply put, it is the ability to recognize, analyze, and question the social and political forces that sustain inequality in society and to take action against them. When we develop a critical consciousness of Generative AI, we acknowledge and address the inequalities the models uphold.

How do we develop a critical consciousness of AI Systems?

We can develop critical consciousness through engaging in three critical components: reflection, motivation, and action. Let’s engage with a short exercise.

Reflection

Within the context of artificial intelligence, reflection answers the question, how does AI create and perpetuate inequality?

Consider GenAI’s impact on:

The environment. GenAI models consume vast amounts of resources, especially water and power, but many global communities lack clean water.
Labor exploitation. GenAI companies often outsource labor to low income countries where workers experience wage theft and poor working conditions.
Data Privacy and Security. GenAI models, especially commercial licenses, lack clear regulation of how data is collected and how data is used.
Social disparities. GenAI models can exacerbate racial, gender, economic, and digital disparities.

Motivation

Within the context of artificial intelligence, motivation answers the question, how capable and committed am I to addressing the inequality AI creates and perpetuates?

Ask yourself:

Does this knowledge affect your understanding or use of GenAI?
How is GenAI’s impact aligned or misaligned with your personal values?
Do you believe you have the ability to address one or more of these impacts?

Action

Within the context of artificial intelligence, action answers the question, what can I do to remove or reduce the inequality AI produces?

Ask yourself:

Which impact motivates me to act?
Do I have any hesitation to act?
What is one small act I can do?

Generative AI and the Challenge of Engagement and Authorship

What is the Challenge of Engagement and Authorship?

Engagement and authorship in academia are two terms that help explain the ways learners interact with learning materials and assessments as they progress through their program. When learners rely on GenAI to either gain or demonstrate knowledge, they risk compromising their own knowledge in some way

How does generative AI compromise engagement and authorship?

Engagement and authorship are connected, but we’ll look at them separately.

Engagement

Engagement in learning is how you gain knowledge or skills. It’s also your desire and curiosity to learn, and your investment in learning. Instructors want you to be actively engaged with the course because they know that deep, meaningful learning can only occur when you are.

Engagement contributes to more meaningful experiences because you:

Learn to think critically about your world—to understand, analyze, and evaluate.
Learn from failure—from your mistakes, miscalculations, and missed opportunities, which are key to deep learning.
Learn through neuroplasticity—your brain’s ability to learn. Your brain makes new neural pathways, removes unused ones, and repairs injured ones.

Authorship

Authorship is best understood as creatorship. We often think of this as academic integrity, but it’s more than just giving credit for ideas or work that is not our own. Meaningful learning requires students to take ownership of and investment in the work they produce.

The ease with which GenAI can create contributes to the challenge of authorship, but here’s what many of your instructors want you to know:

GenAI gets it wrong, a LOT. If it doesn’t know something, it will just make it up or do the best it can. While this might make for some funny memes or videos, AI hallucination compromises the integrity of your work.
GenAI only knows what it’s taught. AI models cannot think. They are trained using data (images, videos, written texts, etc.). And that data has been used without express permission. Suno can generate a song; FireFly can generate an image; Nano Banana can generate a video; but every single AI generator has used the intellectual property of human creators.
GenAI is not smarter than you. It can be a tool to support your learning, but it cannot replace your learning. Your own creativity and intellectual process is much more important and valuable, not just for you, but also for your instructor.