ÌÇÐÄvlog

What is Artificial Intelligence?

Artificial intelligence (AI) is a technology that can perform complex tasks such as learning, problem-solving, and decision making. Generative AI is a type of AI that creates content like text, images, audio, and more. Broadly, Large Language Models (LLMs) are a type of generative AI that generates text. Examples of LLMs that you may have heard of include ChatGPT, Claude, and Gemini. Although AI, generative AI, and LLM are often used interchangeably, it is always good to be specific in your language. This guide will focus on generative AI. 

When you ask a generative AI tool, like an LLM, a question or to perform a task, it processes your input (your prompt) and generates an output based on patterns it learned from the huge amounts of data it was trained on. It is important to understand that LLMs do not “know” things in the way humans do. Rather, they simply predict what text should come next based on statistical patterns in their training data. As such, the quality, quantity, and diversity of training data determines how well the LLM performs. This is why AI companies are constantly seeking more and better training data to improve their model’s performance and capabilities. 

Today’s leading models can complete multi-step tasks with minimal human input, including drafting and organizing files, writing and executing code, and scheduling and managing calendars. This is called “agentic AI.” While these capabilities make generative AI more powerful, they also highlight the importance of thinking carefully about whether and how you choose to use these tools.

The following sections focus on potential practical applications of generative AI for teaching and research. This guide includes an introduction to ÌÇÐÄvlog’s secure AI platform, important ethical considerations, examples of how you can use generative AI tools to support your teaching and research, tips for getting started, and resources for continued learning. 

What is LibreChat?

LibreChat is an open-source, self-hosted AI chat platform that provides the college community with secure access to multiple large language models (LLMs) while maintaining full institutional control over data privacy and security. Unlike commercial AI tools, LibreChat operates within the College's private cloud infrastructure, ensuring that user interactions and data remain completely within our institutional boundaries and are never shared with external entities.

Key Benefits:

• Access to multiple LLMs with the same capabilities as commercial providers
• Enhanced privacy and security within our institutional infrastructure
• Safe integration of sensitive data in compliance with college data classification policies
• Full institutional control over all user data and interactions

Access: Available to all faculty and staff; student access granted upon completion of the course in Canvas.

How to Access: For detailed login instructions and setup guidance, see IT's guide.

Sensitive Data Use: Permitted due to the platform's secure, private environment.

Critical Considerations

It is important to consider not just how you can use AI, but if you should in a specific context considering the things below. These considerations may lead you to use AI tools differently in different circumstances, or to decide to not use them at all for certain tasks.

• Environmental Impact
  • Data centers built to handle the computational demands of generative AI
  • Generating visual content than generating text
  • Consider whether it is necessary to use AI for a task
• Data Privacy
  • Be mindful of what you share with AI
  • Often generative AI settings are by default not private, and the tool will likely learn from your prompts
  • Different tools have different levels of data privacy and security
  • For sensitive data, consider using ÌÇÐÄvlog’s LibreChat platform, which operates within the College’s secure infrastructure 
• Bias from Outputs
  • Generative AI can perpetuate biases present in training data, which, for many LLMs, is all publicly available content on the internet  
  • Evaluate outputs for assumptions, gaps, or perspectives that may be missing or underrepresented 
• Hallucinations 
  • LLMs are statistical models that predict the next word based on patterns in training data, which can lead to false information or citations that appear credible
  • Always validate AI outputs, including information and sources, using reliable references 
•  Transparency
  • Be open about your generative AI use with students, colleagues, and in your scholarly work
  • Disclose your use of AI in course materials: read about the consequences of undisclosed AI use by professors in The New York Times article “”
  • Some citation styles, like , , and , provide guidance on how to cite generative AI. If you have citation questions, reach out to in the library 
• Human Oversight
  • Be deliberate about what permissions you grant AI tools, and keep yourself in the loop on consequential decisions
  • Consider whether a task is too high-stakes to delegate to an AI agent 

What Can Generative AI Do for You?

The following sections describe ways generative AI can support your teaching and research. Rather than recommending specific tools (which rapidly change) we focus on use cases and example prompts. For current tools available at ÌÇÐÄvlog, see the Generative AI tools page.

Teaching Applications

Always carefully review and adapt AI-generated materials to ensure they meet your pedagogical needs and standards.

Course Design and Preparation 

LLMs can serve as thought partners to help make course planning more efficient. 

• Practical Applications: 
  • Assignment Design: Create assignment prompts and criteria for evaluation
  • Generate Case Studies and Examples: Develop relevant examples to illustrate key concepts 
  • Create Study Guides: Generate review materials tailored to your content
  • Rubric Development: Build and customize evaluation criteria 
  • Simulate Student Questions: Anticipate what students might ask to prepare better responses 
• Example Prompts: 
  • “Generate a realistic case study about [topic] that illustrates important ideas about [key concepts].”
  • Upload reading materials: “I’m teaching [topic] to [course level] students. What questions might they ask about [specific concept or part in reading]? Include both clarifying questions and challenging questions.” 

Research and Content Curation 

LLMs, as well as other AI-enhanced academic search engines, can act as your research assistant to find content.

• Practical Applications: 
  • Literature Discovery: Find recent scholarly articles about a specific topic
  • Multimedia Resource Discovery: Find videos, podcasts, websites, etc. to supplement readings
• Example Prompts:
  • Upload readings: “Suggest 3 multimedia resources (videos, podcast, websites) that could supplement these readings.”

Supporting Diverse Learning Styles

LLMs can help you create multiple pathways to the same learning goals.

• Practical Applications: 
  • Multiple Explanations: Create different explanations of the same concept 
  • Accessible Content: Adapt materials for different learning needs and backgrounds
• Example Prompts:
  • “Explain [concept] in three different ways: [e.g. using real-world examples, using analogies].”
  • “Explain [concept] for students with different preparation levels: one explanation for students new to [area], and one for students with some background in [area].”

Visual Content Creation

Generative AI tools can create custom content such as images, presentations, and videos. 

Environmental Awareness Reminder: Visual content creation using generative AI is significantly more energy-intensive than text generation. 

• Practical Applications: 
  • Course Material Illustrations: Create custom content to illustrate concepts, events, scenes, etc.
  • Presentation Enhancement: Generate slide designs, layouts and visuals
• Example Prompts: 
  • “Create a slide template for a presentation about [topic] that includes [design elements]. The style should be [e.g. professional, minimalist, creative].”

Research Applications

Generative AI Research tools can provide a starting point, but they are not replacements for your research expertise.

Privacy Note: There is no guarantee that papers you upload to generative AI tools will not be used to further train AI models. Review each tool’s privacy policy and data handling practices before uploading sensitive material.

Literature Review and Finding Papers 

Generative AI can help you find papers based on your research question or papers you upload. 

• Practical Applications: 
  • Literature Searches: Find relevant papers across databases
  • Gap Identification: Discover under-researched areas
  • Interdisciplinary Connections: Find research from adjacent fields to inform your work 
• Example Prompts:
  • “Help me find recent scholarship on [topic] published in the last [time period]. Focus on work that examines [specific aspect].”

Document Analysis and Synthesis 

AI-powered tools can help you analyze papers and synthesize information.

• Practical Applications:
  • Comparative Analysis: compare multiple sources, identifying themes and contradictions 
  • Content Summarization: summaries of complex academic works to determine if the work is relevant 
• Example Prompts:
  • Upload multiple papers: “Compare these sources on [topic]. Identify any areas of agreement, disagreement, and any gaps in the current scholarly conversation.”

Getting Started

Experiment with different tools and see what works for you. For a current list of AI tools available at ÌÇÐÄvlog, including access levels and whether they permit sensitive data use, see the Generative AI tools page. 

Tips for Using Generative AI

• Be specific in your prompting 

• Provide context about your audience, the purpose, etc.

• Be specific about the format you want your results to take (bulleted list, paragraphs, etc.)

• Iterate your prompts: use your critical thinking to ask models to go deeper, or try another approach

• Try different models as each has different strengths
  • Use multiple tools for research as different tools have access to different databases 
• Break complex tasks into smaller steps

• Use AI to assess outputs: try having one AI evaluate another AI’s work, or ask an AI to critique its own response

• Fact check: AI can make things up (hallucinate)

• • Always verify sources and citations as these tools can make mistakes with academic sources in particular 

Other Resources

ÌÇÐÄvlog Resources 

• The Hastings Initiative’s AI Glossary: From Basic to Technical Terms 
• ÌÇÐÄvlog Library’s
  • Provides basic information about AI like definitions in the Getting Started Tab and links to other readings and resources 
• ÌÇÐÄvlog’s Ethical Considerations and Guidelines
  • Basic overview of ethical considerations and institutional guidelines under Ethical Considerations and Guidelines page
• IT’s Guidelines and Starting Points for Use of Generative AI tools
  • Overview list of Institutional Guidelines and Important Considerations when using these tools 

External Resources 

• • Includes ideas about how faculty can design assignments that involve students using AI in different contexts 
  • Examples of assignments also can be good for thinking more critically about why they’re using AI in the assignment. Is “just for fun” a good reason considering the environmental impacts, for example? 
• • Guidance for educators and administrators on generative AI and institutional policies, teaching, assessments, and more
• Harvard Graduate School of Education Creative Computing Lab’s guide
  • Gives many examples of how generative AI can be implemented through different phases of a project