Skip to Main Content

Research Help

What's the Deal With AI?!

 

 

Let's Break Down the Definitions

 

 

 

 

There are two types of artificial intelligence (AI) that are being widely used in both day to day activities and in higher education - discriminative AI and generative AI (GenAI)

When it comes to discriminative AI, think of these tools as "administrative assistants". They work off of predictive models that distinguish between different classes of data. The tool itself will vary in user friendliness and the way that it is designed, but some of the common uses for this type of AI are:

  • Image and text classification​
  • Speech recognition ​
  • Sentiment analysis 

Real World Example: Gmail uses discriminative algorithms to automatically filter spam and non-spam emails. Ever wonder how your Gmail inbox immediately knows to move your 25% discount on your next pair of shoes email to the "Promotions" section? It's AI!

Generative AI is the more popular set of tools that have taken the information landscape by storm. Chat GPT, which has become a household name is a type of GenAI. These tools can also take on many shapes and forms to, for example, create images, videos, detect anomalies or themes in data, and more. 

When we think of the most used type of GenAI, AI chat bots are likely to come to mind. AI bots can also be called Large Language Models (LLMs) and these models work as such:

  • The user will ask a question or ask the AI bot to complete a task. 
  • The AI tool will answer the user using a method called text prediction, which results in answers that sound more like natural human conversation.​
  • Text prediction will be dependent on what data the AI tool was pre-trained on and how it is being fine-tuned through "reinforcement learning from human feedback”​

Real World Example: You sign-in to ChatGPT to ask it to "create a week long meal prep plan that I can cook in 45 minutes or less, and include a list of groceries I have to shop for". Within seconds, the AI bot will spit out a plan that you can print out or save to your phone. 

Interested in more information about how LLMs work? Watch the YouTube video below!

AI - The Good, The Bad, and The Ugly

Whether we like Artificial Intelligence of not... it is here to stay!

When deciding if AI tools are something you want to integrate into your life and research at Erikson, it is important to look at AI with a holistic eye. These source lists are not meant to persuade you in one way or another, they are meant to empower you to make an informed decision after taking into account all sides of the issue. 

Explore the sources below

Chat GPT, a product of the company Open AI, has a variety large language models (LLM) available to users depending on the account type that one is subscribed to. For example, GPT-3.5 is built on large amount of training material found on the internet from pre-2021 and not connected to the internet.

The latest version of Chat GPT which is GPT 4, claims to be trained on data up to June 2024 and is connected to the internet. The amount of data that the free version has access to will also differ from subscription based Open AI LLMs, that have access to tools like "deep learning". 

Please note that while this word of caution applies to all LLMs, Chat GPT might not be trained on and have access to the most current information available. This can skew the answers that it gives and affect how you decide to use this AI bot. 

AI and Academic Integrity

STUDENTS: On the use of AI and Academic Integrity as defined by the Erikson Student Handbook, (p. 46). 

Academic dishonesty or misconduct can include, but is not necessarily limited to, the following unacceptable conduct:

  • Plagiarism

    • Misappropriating, using or copying of ideas, words or material from another source, including published or posted works of others, without proper attribution. This includes all scholarly and non-scholarly sources of information including, but not limited to books, articles, content from any third party sources including websites, AI sites, etc.)

    • Paraphrasing the source so closely that it barely reads differently than the original source (with or without proper attribution) 

    • Using AI tools in a classroom, or for coursework or assignments without first; 1) confirming with each instructor that AI tools are permitted and 2) accurately and properly citing data/information generated from an AI source or tool, to the extent such data/information is in any way integrated into an assignment, either word for word or through paraphrasing (see https://apastyle.apa.org/blog/how-to-cite-chatgpt for guidelines on how to cite tools like ChatGPT in APA format)

Why Communication Is Key in All Relationships | Lifeworks

It is at the discretion of each individual instructor at Erikson, to determine if AI tools are allowed for classwork and to what extent. Please check in with each of your faculty members to determine what the guidelines around AI tools are. Typically, instructors will have an "AI Disclosure" or "AI Statement" in the course syllabi. If there are no guidelines around use of AI present in your syllabi, do not hesitate to reach out to your instructor for clarification.

email libraryhelp@erikson.edu if you have questions about AI tools and how to ethically use them during grad school.

Remember, communication is Key!

Citation and Attribution with AI Tools

Please Note: You should always check with your instructor before using AI for coursework. 

Source URL: https://apastyle.apa.org/blog/how-to-cite-chatgpt

As with all things related to AI, the norms and conventions for citing AI-generated content are likely to evolve over the next few years. For now, the American Psychological Association (APA) has released preliminary guidelines. Individual publishers may have their own guidance on citing AI-generated content. 

Here are some fundamental ideas that hold true for citing AI generated content, no matter which citation style you're using:

  • Do cite or acknowledge the outputs of generative AI tools when you use them in your work.
    • This includes direct quotations and paraphrasing, as well as using the tool for tasks like editing, translating, idea generation, and data processing. 
  • Do not use sources that are cited by AI tools without reading those sources yourself. There are two different reasons for this:
    • Generative AI tools can create fake citations.
    • These tools may cite a real piece of writing, but the cited content may be inaccurate. 
  • Be flexible in your approach to citing AI-generated content 
    • Emerging guidelines will always lag behind the current state of technology, and the way that technology is applied. If you are unsure of how to cite something, include a note in your text that describes how you used a certain tool. 
  • When in doubt, remember that we cite sources for two primary purposes:
    • first, to give credit to the author or creator;
    • and second, to help others locate the sources you used in your research. Use these two concepts to help make decisions about using and citing AI-generated content. 

Released April 7, 2023

When you cite AI-generated content using APA style, you should treat that content as the output of an algorithm, with the author of the content being the company or organization that created the model. For example, when citing ChatGPT, the author would be "OpenAI", the company that created ChatGPT. 

Here are some guidelines for referencing AI-generated content in APA style:

  • When you reference this content directly in your text, you should include an in-text citation, and an associated entry in your reference list. 
  • If you have used AI tools for some part of your research, you should describe that use in your introduction or methods section, and include the prompts that you used (if applicable). 
  • When referencing shorter passages of text, you can include that text directly in your paper. You might also include an appendix or link to an online supplement that includes the full text of long responses from a generative AI tool. 

Format:

Author. (Date). Name of tool (Version of tool) [Large language model]. URL

Example:

OpenAI. (2025). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat

In-Text Citation Example:

(OpenAI, 2025)

AI Resources

Helpful AI (and FREE!) Tools
  • Questions to ask yourself BEFORE using AI:
    • How much do I want to rely on using an AI tool for a particular project/assignment/task? How much do I want it to "speak" for me?
    • What are the ethical and environmental implications of me using this AI tool? Do the pros of using these tools outweigh the cons?
    • What are tasks I can give AI to make life easier for me? What are tasks that I should reach out for human support on?
    • Are there any data privacy concerns with a particular AI tool? How can I best protect my personal information?

 

AI Tool

 

Description

 

Do I Need an Account to Use?

 

Ethical Use Examples

 

Chat.PDF

ChatPDF is an AI-powered tool that lets you chat with PDFs—upload a document, and it instantly answers questions, summarizes content, or extracts key info using AI. You do NOT need an account, but you can create one if you want to save your chat history. 
  • Explain a difficult concept within the article
  • Break down research data
  • Reword the analysis or conclusion of data

 

Deepseek.AI
DeepSeek AI is a smart conversational chatbot. It uses the DeepSeek-V3 open source large language model (LLM). The AI chat also supports PDF, image, CSV, and Word attachments. They state that multilingual understanding is one of their priorities.   You do need to register for an account with an email, but it is free once you are signed in.
  • Brainstorming research topics
  • Editing and grammar
  • Translation
  • Creating titles or section headers for papers

 

Duck AI Chat

DuckDuckGo AI Chat is a privacy-focused AI chatbot integrated into the DuckDuckGo search engine. It allows users to ask questions and get AI-generated answers while maintaining anonymity—no login or personal data tracking. You can also choose from several LLM models.  You do NOT need an account, you just need to agree to the terms of service
  • Getting feedback on a draft
  • Citation support
  • Creating an outline for a project

 

Goblin.Tools

Goblin.tools is a collection of free, AI-powered tools designed to help with executive dysfunction, particularly for neurodivergent users. It’s privacy-focused and built to reduce overwhelm. You do NOT need an account
  • Breaking down tasks
  • Editing an email or paragraph
  • Weighing pros and cons of an idea

 

Inciteful.xyz

Inciteful.xyz is an AI-powered research tool designed to help academics and writers discover and analyze scholarly papers quickly. It uses networks of citations to:
  • Find key papers on any topic.
  • Visualize connections between research.
  • Speed up literature reviews.
Free to use, with a focus on simplifying complex research.
You do NOT need an account
  • Literature searching and discovery

 

Microsoft CoPilot

Microsoft Copilot is an AI-powered assistant integrated into Microsoft 365 (Word, Excel, Outlook, etc.). It helps with tasks like drafting documents, analyzing data, summarizing emails, and generating ideas—using OpenAI’s GPT-4. You do need to be logged in to your Office 365 account, but the Copilot chat is a free app within your Erikson account
  • Getting background information on a topic
  • Organizing your thoughts
  • Decoding acronyms 
  • Writing website code out for you. 

 

Microsoft Edge Read Aloud (text-to-speech)

Microsoft Edge Read Aloud is a built-in text-to-speech (TTS) feature that reads web pages, PDFs, and eBooks aloud with natural-sounding AI voices. Free to use and no add-ons needed.  You do NOT need an account, but it only works if you are using the Microsoft Edge browser. It is built-in!
  • Reading web pages aloud for accessibility

 

Research Rabbit AI

Research Rabbit is an AI-powered tool designed to accelerate academic research. It helps users:
  • Discover relevant scholarly papers through smart recommendations.
  • Visualize citation networks to explore connections between studies.
  • Organize and track research in collaborative collections.
Free to use and ideal for students, researchers, and writers.
 
You do need to register for an account with an email, but it is free once you are signed in
  • Literature searching and discovery
Let's Talk About AI Detection Tools:

AI Detectors...Are AI! AI detectors are usually based off similar LLM models of the tools they are attempting to detect​. Imagine that the AI detector is asking, “Is this the sort of thing that I would have written?”​

AI Detectors look for two things in text:​

  • Perplexity – how unpredictable is the text?
    • AI flagged: If perplexity is low, then the text will be more predictable and easy to read, but it also might sound more monotonous. Typically low perplexity text will be flagged as AI written text.
    • Human flagged: If perplexity is high, then the writing might sound more creative but it will usually have more grammar mistakes and issues with content flow. This type of text will be not be flagged as AI written, therefore the detector will determine a human wrote it. ​
  • Burstiness – how much variation is there in sentence structure and length? ​
    • AI Flagged: If burstiness is low, then it will be flagged as AI written text, for similar reasons to perplexity. 
    • Human flagged: If burstiness is high, meaning the text is more unpredictable, the AI detector will not flag it as AI text; it will be considered human written content. 

List of Most Common AI Detectors in Higher Ed: ​

  • Turnitin AI detection tool​
  • Compilatio​
  • GPTZero​
  • PlagiarismCheck​
  • DetectGPT​
  • Quillbot AI Detector​
  • Grammarly Authorship​

Challenges with Using AI Detection Tools 

Weber-Wulff, D., Anohina-Naumeca, A., Bjelobaba, S. et al. Testing of detection tools for AI-generated text. Int J Educ Integr 19, 26 (2023). https://doi.org/10.1007/s40979-023-00146-z

Taking a look at the research above by Weber-Wulff, et al. (2023), 14 tools (including the Turnitin AI Detector) were studied for accuracy and multiple error types (including false positives)​. No AI detector, including Turnitin, scored 80 percent or above. These tools have also affected ESL learners and international students at an alarming and disproportionate rate. As great as these tools sound on the surface, they unfortunately are not where they need to be to prove with 100 percent certainty that a piece of text was written by AI or not. The advancement of Generative AI as a whole is moving too quickly for detectors to keep up. Below are some statements from Universities that have decided to disable AI detectors at their institutions due to their faulty nature:

SO, how do we combat the overreliance of AI usage and issues of AI and academic integrity?!

There is not a perfect answer or solution to this issue. There are many ways to think about staying ahead of the curve of AI in the classroom, such as:

  • Ensuring you have a thorough AI statement in all course syllabi on how AI is or is not permitted for course work
  • Changing assignments from the "traditional paper" to more active options such as:
    • classroom discussions
    • narrated presentations
    • infographics
    • podcasts
    • thesis defenses 
    • integrating AI into the assignment and having students cite and explain how AI was used 
    • Turning on editing history in Word or Google docs to track the version history of student assignments
    • etc.

If you decide to use the Turnitin AI detection tool, make sure that it is just ONE of MANY tools that you use to determine if AI was used in student work. Need more help? Reach out to the librarians via libraryhelp@erikson.edu or book a research consultation with us!

Best Practices to secure and preserve your data privacy when using Generative AI and LLMs: 
  • Clear Consent:
    • Before utilizing an AI tool, become familiar with consent mechanisms in place. The "settings" or "account" section of an AI tool should provide the app's terms of use and privacy policy. Typically these tools also have options for enabling/disabling your data being used to improve the model, options to connect/disconnect external applications, and options to delete any stored history. Its important to know how your data will be used and for what purpose.
  • Limit Data Retention:
    • Set clear policies on how long data will be retained. Once its utility has been exhausted, securely delete or anonymize it.
  • Data Anonymization:
    • To protect your privacy and that of others, do not input any identifying information into any AI tool, or information that is not publicly available on the internet.
    • STAFF/FACULTY - On Data Privacy and the Erikson HR Policy Manual: Although Erikson does not have specific guidelines around AI use for faculty and staff, please consult section III.7 (Confidentiality) of the Erikson HR Policy Manual and the FERPA section of the U.S. Department of Education website for more details on protecting different types of information as an Erikson employee.

Interesting AI Reads