worldcat
GUILLERMO AND MICHÈLE DE LA DEHESA LIBRARY CATALOG
ADVANCED SEARCH EXPERT SEARCH HOW TO SEARCH

AI Basics

What is Artificial Intelligence?

Artificial Intelligence (AI) refers to the ability of computer systems to perform tasks that normally require human intelligence, such as reasoning, learning, problem-solving, and decision-making. By analyzing large amounts of data, AI systems can identify patterns, generate insights, and support predictions or automated decisions.

Generative AI is a branch of artificial intelligence that focuses on creating new content in response to user instructions or prompts. These systems can generate text, images, audio, video, code, and other types of content by learning patterns from large datasets during training. As a result, generative AI tools have the potential to transform content creation, research, education, and many other fields by making the production of information faster and more accessible.

However, generative AI also presents important challenges. Its outputs may contain inaccuracies, reflect biases present in training data, or reproduce copyrighted material without clear attribution. For this reason, AI-generated content should be critically evaluated and used responsibly, particularly in academic, professional, and research contexts.

AI Issues

Read about AI Issues

The following sections summarize some of the main ethical, technical, and societal challenges associated with the use of Artificial Intelligence. Understanding these issues is essential for using AI responsibly in academic, professional, and everyday contexts.

1. Ethical Considerations
  • Bias and fairness: AI systems can reflect or amplify biases present in their training data, potentially leading to unfair or discriminatory outcomes.
  • Misinformation: Generative AI can create convincing but inaccurate content, including fake news, deepfakes, and misleading information.
  • Intellectual property: The use of AI-generated content raises questions around authorship, copyright, attribution, and the use of copyrighted materials in model training.
  • Accountability: Determining responsibility for AI-generated outputs can be complex, particularly when errors or harmful content occur.
  • Access and equity: Unequal access to AI technologies may contribute to existing educational, economic, and social inequalities.
2. Quality and Reliability
  • Accuracy: AI-generated content may contain inaccuracies, outdated information, or fabricated details.
  • Consistency: Responses can vary and may not always be reliable, even when given similar prompts.
  • Depth and originality: While AI can generate content quickly, it may lack nuanced understanding, critical thinking, or genuine creativity.
  • Model degradation: As AI-generated content increasingly becomes part of online training data, concerns have emerged about declining output quality over time.
3. Privacy and Security
  • Data privacy: AI systems may process large amounts of user data, creating potential risks around confidentiality and data protection.
  • Information exposureSensitive information included in training data or user interactions could be inadvertently disclosed or misused.
4. Environmental Impact
  • Resource consumption: Training and operating large AI models requires substantial computing power, resulting in significant energy use and environmental costs.
5. Human Impact
  • Overreliance on AI: Excessive dependence on AI tools may reduce opportunities to develop critical thinking, creativity, and other essential skills.
  • Workforce disruption Automation may transform or replace certain roles and tasks across industries.
  • Hidden laborAI systems rely on large numbers of human workers for data labeling, moderation, and quality assurance, often under challenging working conditions.

How can I use AI?

AI in Research

AI tools can be used to support different aspects of the research process:

  • Hypotheses Generation: The use of AI allows researchers to generate research questions based on a given dataset or topic. The result can be used as a starting point for researchers to refine and develop into working hypotheses.
  • Literature Review: Researchers can use AI to support their research process, allowing for a faster performance during the literature review process by analyzing and summarizing a body of literature on a topic, as well as identifying relevant trends, patterns, and gaps in existing knowledge.
  • Data Analysis: AI can provide support when processing and analyzing large datasets, allowing for a greater ease identifying emerging trends, correlations, outliers, and other patterns.
  • Experiment Design: The use of AI algorithms can support researchers when designing experiments by suggesting variables, methodologies, and potential outcomes based on the implemented data.
  • Communication of Findings: AI can act as a support drafting, proof-reading, and editing research papers.
  • Collaboration and Networking: Implementing an AI-driven recommendation system could facilitate researchers connect with peers, collaborators, and experts in their field, activities that foster collaboration in interdisciplinary spaces.
AI in Publishing

Since the use of AI has become widespread, its role in academic publishing has been a central point of discussion in scholarly circles. The most relevant topics surrounding those debates have been:

  • Authorship and Attribution: The use of AI in research and writing has incited the debate of how to assign authorship and attribution, since AI systems have been known to generate portions of academic articles, opening discussions about whether to considerate AI as an author or a tool.

  • Plagiarism and Originality: AI-generated content may increase the risk of plagiarism if it is used without proper acknowledgment or critical review. Researchers and students should follow institutional guidelines on citation and responsible AI use.

  • Intellectual Property: Questions remain about the ownership of AI-generated content and the use of copyrighted materials in AI systems. Understanding licensing, copyright, and intellectual property rights is essential when using AI in academic work.

  • Quality and Reliability: AI-generated outputs should be carefully reviewed for accuracy, originality, and relevance. Human oversight remains essential to ensure the quality of academic research and publications.

  • Ethics and Transparency: Responsible AI use requires transparency about how AI has contributed to research, writing, or content creation. Users should also be aware of potential biases, inaccuracies, and ethical implications associated with AI-generated content.

AI Tools

In February 2025, IE University entered into a strategic partnership with OpenAI, becoming one of the first universities worldwide to do so. As a result, ChatGPT Edu has been adopted as the University’s standard AI platform for general use.

To learn more about ChatGPT training opportunities and how to access a complimentary license, please visit the AI Training section.

For faculty members, it is also possible to obtain individual licenses by participating in a co-teaching certificate program or making a request through your platform.

Tool Source / How to Use Information Accuracy Key Strengths Limitations Best For
Microsoft Copilot Context-aware responses with improved reasoning.
  • Strong contextual understanding and advanced reasoning.
  • Designed to support productivity across Microsoft applications and workflows.
  • Supports file analysis, presentation creation, PowerPoint assistance, and other workplace tasks.
  • Offers flexible interaction styles, ranging from creative ideation to precise execution.
  • Some organizations may have privacy concerns due to deep Microsoft integration.
  • Usage limits may apply depending on the subscription plan.
 Organizations and professionals heavily invested in Microsoft tools; productivity, documentation, collaboration, and enterprise workflows.
Perplexity Provides source citations by default, making answers easier to verify.
  • Provides source citations by default.
  • Delivers structured, research-oriented responses.
  • Excels at deep web search and information synthesis.
  • Offers advanced article summarization and long-form content analysis.
  • Native image generation capabilities are still unavailable.
  • Some multimodal features remain under development.
 Academic research, source-confirmed search, and deep information discovery.
ChatGPT Broad knowledge base combined with strong reasoning capabilities; maintains continuity across conversations, projects, and uploaded materials.
  • Broad knowledge base and high versatility.
  • Maintains continuity across conversations and projects.
  • Supports voice interactions, branching conversations, and project workflows.
  • Includes Advanced Voice Mode, model selection, file analysis, and collaboration features.
  • Balances creativity, logical reasoning, and multimodal capabilities.
  • Responses can occasionally be longer than necessary.
  • Complex reasoning tasks may sometimes produce inconsistent outcomes.
 Creative and logical problem solving, document analysis, and project workflows.
Claude (Anthropic) Efficient reasoning with strong continuity across long conversations and projects.
  • Strong contextual understanding and efficient reasoning.
  • Designed with transparency and collaboration in mind.
  • Supports project exports, web browsing, PDF analysis, and coding assistance.
  • Recognized for software development and knowledge work.
  • Advanced memory features are primarily available through paid plans.
  • Some file creation and manipulation capabilities are restricted by security safeguards.
 Software development, coding tasks, and enterprise AI workflows.
Tool Source / How to Use Purpose Key Features Best For
Scite Uses AI-powered citation analysis to help researchers evaluate how scientific papers have been cited and discussed within the academic literature.
  • Advanced citation analysis.
  • Integration with academic databases.
 Literature reviews, research finding evaluation, and citation-based analysis.
Elicit Automates empirical research workflows and literature discovery.
  • Displays relevant papers.
  • Summarizes key information.
  • Supports systematic reviews and evidence extraction.
 Empirical research, systematic reviews, and accelerating evidence gathering and analysis.
Inciteful Supports research discovery by revealing citation relationships and identifying influential papers within a research field.
  • Citation network analysis.
  • Discovery of related and highly connected research articles.
  • Identification of influential and emerging publications.
 Exploring citation networks, identifying foundational research, and discovering related literature.
Research Rabbit Helps researchers explore academic literature through visual maps and personalized recommendations.
  • Interactive visualizations of research topics.
  • Personalized article recommendations.
  • Continuous discovery of relevant literature.
 Literature visualization and personalized exploration.
AI PROMPTS

Prompt engineering is the practice of conceptualizing, defining and perfecting the inputs fed to AI language models in order to achieve the desired outputs required by the user.

It involves designing prompts that steer the model in generating responses that are accurate, relevant, and aligned with the user’s goals.

In an academic setting, mastering prompt engineering allows students and researchers to unlock the full potential of AI in research and development of diverse academic support features and activities.

Read some tips and the CLEAR Framework for Effective Prompt Writing
The CLEAR Framework for Effective Prompt Writing
The CLEAR Framework is a simple approach to writing effective prompts that help AI tools generate more accurate, relevant, and useful responses. The framework stands for Concise, Logical, Explicit, Adaptive, and Reflective, providing a structured method for improving interactions with AI.
  • Concise Keep prompts focused and avoid unnecessary information. Clear and direct instructions help AI understand the task more effectively.
    Example:
    Instead of "Write a detailed essay about the ethics of AI how people use them and what they can do about them" use "Summarize the main issues and ethics of the use of AI and display them as a list."
  • Logical Structure prompts in a clear sequence, outlining the role, task, requirements, and desired output.
    Example:
    "As an climate change expert, explain four ways climate change is affecting desertified areas and suggest a mitigation strategy for each of them in bullet points."
  • Explicit Clearly define the purpose, audience, tone, format, and length of the response.
    Example:
    "Write a 300-word summary of desertification causes for a high school audience using an informative tone."
  • Adaptive Refine prompts based on the responses you receive. Small adjustments can significantly improve the quality and relevance of the output.
    Example:
    If a response is too technical, revise the prompt to request "a simple explanation suitable for beginners."
  • Reflective Review AI-generated content critically for accuracy, completeness, relevance, and potential bias. Use these observations to improve future prompts.
    Example:
    If important information is missing, revise the prompt to request additional details or address specific gaps.
Tips for Writing Effective AI Prompts
  • Be Clear and Specific The quality of an AI response depends heavily on the quality of the prompt. Clearly state what you want, including the topic, format, audience, length, or level of detail. The more precise the output, the better that GenAI can tailor its answer. Make sure not to make your prompt too long, since very long or detailed prompts can overwhelm GenAI models and lead to inexact outputs.
    Example:
    Instead of "Tell me about desertification," try "Summarize the climate change factors contributing to desertification in a 300-word essay."
  • Provide Examples Examples help AI understand the style, tone, and level of detail you expect.
    Example:
    "Using the same academic style as this optometrist webpage, create a guide for eye care tips."
  • Choose the Right Type of Prompt Different goals require different prompting approaches.
    Open-ended prompts encourage exploration and brainstorming.
    "What are the main Greek Pantheorn deities?"
    Closed-ended prompts request specific, structured information.
    "Identify all the twelve Olympic gods of the Greek Pantheon and explain their characteristics in two to three sentences"
  • Assign a Role or Persona Giving the AI a role can help generate more relevant and targeted responses.
    Example:
    "Act as a university professor in economics and explain inflation to first-year students."
  • Refine Through Follow-Up Prompts Effective prompting is often an iterative process. If the first response does not meet your needs, ask follow-up questions, request revisions, or provide additional context.
  • Start a New Conversation When Needed If a conversation becomes unfocused, consider starting a new chat. You can summarize the useful parts of the previous discussion and use that summary to guide the new interaction.

Further Resources:

Google: Gemini for Google Workspace – Prompting Guide 101

A detailed guide from Google introducing best practices for writing effective prompts. It can be applied to all GenAI tools.

 

OpenAI: Prompt engineering best practices for ChatGPT

An outline of best practices and prompt engineering instructions for ChatGPT, created by OpenAI. It includes tips on structuring prompts, giving clear instructions, and iterating to improve responses.

 

OpenAI Related resources from around the web

Lists of prompting libraries & tools, prompting guides, video courses and papers on advanced prompting to improve reasoning compiled by OpenAI.

How to cite AI

Quick Reference Guide to Citing AI Tools

Important!

Like AI, citation requirements for AI-generated content are in constant evolution. To ensure your use of AI is correctly disclosed, make sure to check the guidelines provided by your instructor, department, publisher, or journal, as requirements may vary by discipline and institution.

Publishers Policies on AI

After the increase of the use of AI in the academic process, different publishers have established their own policies regarding permissible use and attribution.

Some of those policies are as follows: