Google Gemini and Gems is approved for use by all faculty and staff and is accessible within PAWS. Students in courses currently piloting Gemini and Gemini Notebook also have access.

What is Google Gemini and Gems?

Google Gemini is a generative artificial intelligence (GenAI) tool developed by Google. Gemini uses a conversational interface that allows users to interact with AI using prompts and can support tasks such as brainstorming, writing, summarizing, planning, analyzing information, coding assistance, and working with uploaded materials.

Gems are customized versions of Gemini that can be configured with instructions for a particular purpose or recurring task. Users can create Gems that follow specific directions, adopt a particular role, use a preferred response format, or reference selected files as contextual knowledge.

For more information on what GenAI is and how it can be used to support teaching and learning, see the GenAI webpage.

What is the purpose of Google Gemini and Gems?

Gemini can serve as a general-purpose AI thinking partner for exploring ideas, generating and revising content, summarizing information, planning activities, and working through problems.

Gems extend these capabilities by allowing users to configure Gemini for a recurring purpose. Rather than providing the same background, instructions, or expectations in every conversation, a user can create a Gem with instructions that Gemini applies each time it is used. For example, an educator could create a Gem that generates discussion questions using the same pedagogical approach and format each week.

Google Gemini and Gems are currently approved for use by faculty and staff. It is accessible via the PAWS channel ‘AI Tools’. (USask NSID login required)

Not what you’re looking for?

Why use Google Gemini and Gems?

Google Gemini and Gems serves as a “thinking partner” to generate ideas, written text, and images.

For teaching and learning, Gemini chat can be effective at generating ideas for learning outcomes, assessments, activities, teaching examples, and more. Gems allow users to save tailored instructions, context, and resources for a particular purpose. In teaching and learning, a Gem might support a recurring task or be designed for specific learning activity, such as guiding students through a case, providing structured practice or prompting reflection.

A useful rule of thumb is:

Gemini = a conversation for the task at hand.
Gem = a customized AI experience designed for a particular purpose, learning activity, or recurring workflow.

  • Google Gemini and Gems references the sources it used when producing its responses.  
  • USask has an enterprise license through Google. Access the tool via PAWS or log in directly using your NSID@usask.ca email address.
  • Your data is protected when logged into the enterprise license.
  • User inputs and Google Gemini and Gems outputs are not used to train GenAI models. 
  • Google does not claim intellectual property rights to either the user input or the Google Gemini and Gems output.

Learning Technology Ecosystem (LTE) Principles

Google Gemini and Gems most directly address these LTE goals:

Accessible - Learning must be found easily at any time, and all learners and teachers have equitable access, regardless of culture, language, ability etc.

Efficient and easy to use- Learners need to work in a system that is fluid and requires a minimum number of steps in systems that are intuitive and integrated.

Best Practices

  • Review the GenAI Literacy Framework to guide appropriate GenAI tool use for you and your students.
  • Complete the USask AI Literacy Tutorial and ask your students to complete it as well to develop a shared understanding of responsible and effective GenAI use.

DO

DON'T

 Use Google Gemini and Gems mindfully by considering the environmental impact of each interaction. GenAI tools like Gemini require significant resources to operate.

 Don't overuse Google Gemini and Gems for tasks that could be done more effectively by other means. Consider whether AI is the best tool for the job before using a GenAI tool.

 Interact with Google Gemini as a thinking partner by prompting it to generate and give feedback on ideas.

 Don't assume Gemini's responses are accurate because they sound confident or plausible.

 Ask Gemini to provide feedback or alternative perspectives rather than automatically accepting its first response.

 Don't use Gemini or Gems as GenAI or plagiarism detectors.

 Use an effective prompting strategy to get the results you're looking for in the least amount of time.

 

 Verify factual claims, references, quotations, and other generated information.

 

 Review generated teaching materials for disciplinary accuracy and appropriateness.

 

 Use Gems for appropriate recurring tasks where consistent instructions would be useful.

 

 Give Gems clear instructions about their purpose, intended audience, constraints, and desired output.

 

 Review files before adding them as knowledge to a Gem.

 

 Be transparent about appropriate GenAI use.

 

 Use Gemini mindfully and consider whether a non-AI tool could accomplish the task effectively.

 

Support for Google Gemini and Gems

Technical Support

(Why isn't this working?)

USask IT supported

Vendor Supported

Training Support

 (How do I learn to use this tool?)

USask IT supported

Gwenna Moss Centre for Teaching and Learning Supported

Vendor Supported
  • Email GMCTL for support with using GenAI to design learning activities and assessments.

Teaching Support

(How do I teach with this tool?)

Gwenna Moss Centre for Teaching and Learning Supported
  • Email GMCTL for support with using this tool to design learning activities, assessments, one-to-one consultations, and related workshops.

Tool Evaluation

Technologies are evaluated based on their alignment with the Learning Technologies Ecosystem Principles.  Click to see explanations of each principle and a justification of the rating. You can also view a complete blank rubric to see more details or read about the assessment process.

Learning must be found easily at any time, and all learners and teachers have equitable access, regardless of culture, language, ability etc. 
More Information

Rating: Moderate Concerns

  • Accessibility standards:
    • Gemini provides a web-based conversational interface and supports text-based interaction. Accessibility of individual Gemini features should be considered separately, particularly where generated content includes images, tables, multimedia, or other formats.

  • Cost of use for USask students:
    • Google Gemini (Basic) is currently only available to USask faculty and staff. Students in courses currently piloting Gemini and Gemini Notebook also have access.

  • Platform/device:
    • Gemini is available through a web browser and through Gemini mobile applications. Some features may differ between the web and mobile versions.

  • Offline Access:
    • Gemini's core GenAI functionality requires internet access.

Learning is a process of meaning-making, constructed through learning with others, and as a part of an intentional, deliberate system within a course and across experiences.
More Information

Rating: Serious Concerns

  • Collaboration:
    • Gemini is primarily designed for interaction between an individual user and an AI system rather than communication and collaborative knowledge construction among learners. Educators can design collaborative learning activities that incorporate Gemini, but collaboration is not a core feature of the tool itself.

  • Sharing:
    • Gemini-generated content can be transferred or shared outside the tool. Users can also share Gems they create, allowing others to use a Gem configured with particular instructions, context, or resources. This may support the reuse of Gems designed for specific learning activities or purposes.

Learning is refined and extended through prompted and supported opportunities to focus on understanding and next steps.
More Information

Rating: Moderate Concerns

  • Reflection and revision:
    • Gemini supports iterative interaction. Users can ask questions, respond to feedback, reconsider ideas, and request revisions during a conversation.

Learning is most effective when systems are designed to help learners find, create, and/or repurpose significant content for the value of themselves and others.
More Information

Rating: Works Well 

  • Creating:
    • Gemini can generate, transform, summarize, reorganize, and revise content in response to prompts. This may support iterative creation and experimentation.

  • File format:
    • Generated content can be copied, transferred, or used in other applications, and some Gemini functionality supports working with files.

Learners create and control spaces for learning, understanding and retaining ownership, and purposefully choosing how and when they share.
More Information

Rating: Moderate Concerns

  • Archiving, saving, and exporting data:
    • Gemini can retain conversations and generated content according to the functionality and settings available to the user's account. Generated material can also be transferred into other applications.

  • Data privacy and ownership:
    • Google does not claim intellectual property rights to either the user input or the Gemini output.
    • Your data is protected when logged into the enterprise license.
    • No history of conversations is kept. User inputs and Gemini outputs are not used to train GenAI models when logged into the enterprise license.

  • Bias:
    • Like other generative AI systems, Gemini can generate content that reflects biases in its training data, design, or the context supplied by users. Bias may not always be obvious within an individual response. Users should critically evaluate outputs, seek alternative perspectives where appropriate, and avoid treating Gemini responses as neutral or authoritative.
  • Sign Up/Sign In:
    • Access via PAWS using your USask Google account.

  • Customization:
    • Gemini can be customized through prompting, and Gems provide additional customization reusable instructions describing the role the Gem should play, the task it should perform, relevant context, and the desired response format and added files.

Learners need to work in a system that is fluid and requires a minimum number of steps in systems that are intuitive and integrated.
More Information

Rating: Works Well

  • Interface:
    • Gemini uses a conversational chatbot interface. Users enter prompts and continue interacting with the tool through follow-up questions and instructions.

  • Additional Downloads:
    • No additional software is required to use the Gemini web application. Mobile applications are also available.

  • Functionality:
    • Core Gemini functionality is available across supported platforms, but some advanced capabilities may differ according to account type, license, device, or application.
  • Environmental Impact:
    • Generative AI systems require substantial computational resources and energy to operate. Use Gemini mindfully and consider whether GenAI is necessary for a task before using it.

Learners exist in accessible networks, and connect to the experiences, concepts, people, and ideas that they need.
More Information

Rating: Moderate Concerns 

  • Scale:
    • N/A

  • Communcation Flexibility:
    • Gemini supports multiple modes of interaction, including textual and, depending on the available Gemini features, audio, visual, and multimodal communication.

  • Representation and Inclusivity:
    • Gemini can generate content representing different perspectives, languages, contexts, and forms of expression. However, AI-generated representations may contain cultural assumptions, stereotypes, omissions, or bias.
    • Users should critically review generated content, particularly when it represents cultures, identities, communities, or perspectives outside their own experience.

Learning and feedback are iterative, and assessment comes from multiple sources, including self, peers, teachers, and outside experts.
More Information

Rating: Serious Concerns

  • Feedback:
    • Gemini can provide iterative AI-generated feedback on ideas and content. Users can respond to that feedback and request clarification or further suggestions. However, Gemini is not designed to store feedback from multiple human participants alongside student work in the manner of an assessment or learning-management system.

  • Sources for Feedback:
    • Gemini can provide AI-generated feedback but does not provide built-in mechanisms through which self, peers, educators, and outside experts can collectively provide feedback on learning.
  •  Learning Data:
    • Gemini is not designed to provide educators with access to student learning data or channels for reviewing individual students' interactions and providing feedback on those interactions.