About this module
Welcome to our module that explores the limitations of Generative AI (genAI). While genAI tools are a fantastic resource in your learning and study toolkit, it’s crucial to recognise they’re not infallible. This module is focused on building your awareness of reliability issues and other limitations that impact genAI output. Limitations that affect your engagement with and content generation in genAI.
Like other tools and information systems, genAI has both strengths and weaknesses. It's ability to generate content or outputs (learn more), from textual responses to creative designs, is underpinned by complex algorithms and vast datasets (see genAI basics). By examining weaknesses like hallucinations, inaccuracies and biases, you can better use genAI in a digitally savvy and fluent way.
Big Question
This module explores the limitations of genAI. Ask yourself 'How much do I trust the content produced by genAI tools?'
Consider how using inaccurate or flawed outputs impacts your decision-making both personally and professionally.
Explore a real-life genAI and trust story by clicking on the arrows to move through the interactive below.
Caution
Generative AI creates content based on the data it's trained on and the way you prompt or ask questions. If it learns from incorrect, outdated or biased information sources, genAI can produce wrong or misleading information.
Purpose of this module
Answering the question of how much do I trust genAI relies on being aware of potential pitfalls in the quality and reliability of generative AI output. Building your awareness of limitations, gaps, biases or other inaccuracies supports your evaluative and critical thinking skills in using these tools.
Learning objectives:
- Recognise that outputs produced by genAI have limitations
- Identify how limitations impact the reliability of genAI output
- Explore different limitations of genAI output
- Use strategies to minimise the risk of genAI limitations
This module will take approximately 25 minutes to complete.
Attribution and acknowledgement
Crediting creators or attributing content is a core part of both academic integrity and of being a digital citizen more broadly. This guide has been informed by the following resources in terms of the using genAI practical concepts, their instructional approach and their Open Education design:
- AI in Education modules created by University of Sydney students and staff
- Artificial Intelligence Digital Essentials created by University of Queensland
Definitions and other written content were augmented or refined through use of text based generative AI (OpenAI, 2023). Images were generated through use of generative AI (Ideogram, 2023).
This module was created by Deakin Library. The text and layout of 'GenAI limitations' module are © Deakin University 2023 and licensed under a CC BY-NC 4.0.