Skip to Main Content

Artificial Intelligence

Guide to Artificial Intelligence (AI) in health sciences research.

What is AI?

Artificial intelligence (AI) is also referred to as machine learning (ML) although they are different. Similarly, Large Language Models (LLMs) are often referred to as AI and fit under the umbrella of AI with ML but neither demonstrates actual intelligence.

AI (Artificial Intelligence) "is the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable [emphasis added]." (McCarthy, n.d.)

Intelligence of created systems and algorithms is typically compared to human intelligence. Sometimes LLMs and ML products can appear to have human intelligence, but it is simply the product of coding, not actual intelligence.

ML (Machine Learning) is "algorithms that give computers the ability to learn from data, and then make predictions and decisions". Examples include automatically detecting spam emails, suggesting videos to watch after finishing one, etc. (CrashCourse, 2017) 

LLMs (Large Language Models) "can generate natural language texts from large amounts of data. Large language models use deep neural networks, such as transformers, to learn from billions or trillions of words, and to produce texts on any topic or domain. Large language models can also perform various natural language tasks, such as classification, summarization, translation, generation, and dialogue." (Maeda & Chaki, 2023)

GPT (Generative Pre-trained Transformer) "models give applications the ability to create human-like text and content (images, music, and more), and answer questions in a conversational manner." (What Is GPT AI?, n.d.)

Uses & Limitations of LLMs

Large Language Models (LLMs)

While some LLMs are more recognizable than others, there are quite a few options available. Here is an incomplete list of LLMs you may want to investigate. (Staines, 2023)

Sampling of LLMs

  • ChatGPT
  • Consensus
  • CORE-GPT
  • DataSeer
  • GDELT Project
  • Hum
  • Iris.ai
  • LASER AI
  • Perplexity.ai
  • Prophy
  • Scholarcy 
  • Scite.ai 
  • SOMA
  • Writefull
ChatGPT

 

Uses and Limitations of LLMs

UW Medicine Office of Research and Graduate Education:

  • OpenAI has a call for proposal on their website right now. UW Principal Investigators (PIs) should not apply for these grants.
  • OpenAI will claim possession and rights of dissemination for all material in the proposal and all background IP.  
  • These terms are not acceptable to UW and many other universities.

Please direct any questions regarding the OpenAI call for proposals to rgeadmin@uw.edu
 

Uses

Limitations

  • It can be a starting place, but do not rely on it for factual information or research.
  • It is not a search engine, but uses vast amounts of data to generate responses that appear to make sense.
  • LLMs and AI are known for producing hallucinations, where the program presents and defends false information as if it were factual.
    • Most LLMs have created citations to defend its statements, but these citations can be entirely fabricated.
  • LLMs have access to knowledge up to a given date. You may ask the LLM when that date is, but it is best to go to the developer's notes, if available, to confirm whether its information is up to date.
  • LLMs are improving, but still provide false citations.
  • LLMs' privacy policies may allow the creators to sell and profit off of your personal information.
  • Submitting a manuscript to a LLM for writing assistance may violate requirements from journals you wish to publish in or your institution.
  • Anything you submit to an LLM may become a part of the LLM's learning corpus.
  • Discussion starter regarding privacy, intellectual property, research integrity, ethical consumption, and more!

Guidance for AI Development and Use

Bias in AI and LLMs

Journal and News Articles on Bias in AI and LLMs

Cost of AI and LLMs

Are LLMs Free?

Classroom Use of AI and LLMs

Using AI in the Classroom

Use Caution

Academic Standards

Rethinking Plagiarism and Cheating

From Ditch That Textbook by Matt Miller.

alt text below in plain text

Alternative Text

It's time to rethink "plagiarism" and "cheating"

As you read through the ordered list, consider these questions:

  • Which of these would you consider "cheating"? 
  • Which of these is relevant to our students' future? 
  • Which of these would you use in your work as an adult? 

Organized on a spectrum from Bot-created (1 on the list below) to Student-created (6 on the list below):

  1. Student plugged prompt into AI, copied response and submitted it to teacher. 
  2. AI created a response. Student read, edited, adjusted, and submitted. 
  3. Student created multiple AI responses, used the best parts, edited, and submitted. 
  4. Student wrote main ideas. AI generated a draft and offered feedback to improve. 
  5. Student consulted internet/AI for ideas, then wrote and submitted. 
  6. Student wrote all assignment content without consulting AI or the internet. 

Ditch That Textbook. Graphic by Matt Miller (@jmattmiller) DitchThatTextbook.com. 

(Miller, 2022)

Citing AI and LLMs

Co-Authorship

Many journals have rules about whether AI and LLMs can be considered a co-author. Investigate the Author Guidelines for any journal in which you are considering publishing before using a LLM to revise or support your work.

Works Cited

Consensus on how or whether to cite AI and LLMs is in progress. Look to your instructor and/or syllabus for specific instructions on whether you may use AI and/or LLMs in your coursework. 

Because LLMs are generating sentences based on others' work, err on the side of caution by citing the LLM used.

Copyright

Learning Forums & Resources

Learn the basics with this 12-minute instructional video by CrashCourse produced in collaboration with PBS Digital Studios:

UW Resources

Library Research Guides

Think Pieces

References

CrashCourse (Director). (2017, November 1). Machine Learning & Artificial Intelligence: Crash Course Computer Science #34. https://www.youtube.com/watch?v=z-EtmaFJieY

Maeda, J., & Chaki, E. (2023, April 3). Concepts Overview for LLM AI. Microsoft Build. https://learn.microsoft.com/en-us/semantic-kernel/concepts-ai/

McCarthy, J. (n.d.). What is AI?: Basic Questions. Professor John McCarthy. Retrieved May 3, 2023, from http://jmc.stanford.edu/artificial-intelligence/what-is-ai/index.html

Miller, M. (2022, December 17). ChatGPT, Chatbots and Artificial Intelligence in Education: How to define "cheating" and "plagiarism" with AI. Ditch That Textbook. Retrieved December 6, 2023, from https://ditchthattextbook.com/ai/

Staines, H., et al. (2023). Charleston In Between: The Future of Scholarly Communication in a ChatGPT World. https://www.charleston-hub.com/the-charleston-conference/welcome/charleston-in-between/

What is GPT AI? - Generative Pre-Trained Transformers Explained - AWS. (n.d.). Amazon Web Services, Inc. Retrieved August 10, 2023, from https://aws.amazon.com/what-is/gpt/