What is genAI?
Generative Artificial Intelligence (commonly referred to as generative AI or genAI) is able to create a range of digital content from user prompts. This technology learns by picking up patterns from large amounts of training data, which it uses to generate content in response to user prompts. It can be anything from human-sounding chatbots to predictive text on your phone.
It's very likely you’ve used a form of generative AI even if you weren’t aware of it.
A brief history of AI
While genAI technology has rapidly expanded in the past year with the launch of ChatGPT in November 2022, it’s important to note that the foundations of this technology have existed for many years prior.
Advances in AI timeline
Drag the sliding button to explore a brief history of key moments in AI from neural networks to ChatGPT.
How does genAI work?
In order to understand genAI, it is important to know how it works. Watch this video clip (6:20) explaining the process, including how tools like ChatGPT work and providing examples of AI in everyday life.
This video clip introduced some key terms relating to genAI, including machine learning, Large Language Models (LLMs) and neural networks. You can revist these terms and many others that you come across in these modules by visiting the genAI Glossary.
Caution
AI tool(s) currently have no concept of fact or accuracy. They simply look for patterns in their training data (learn more), then produce what they think is the next most likely word or phrase.
What can genAI tools do?
There are many genAI tools out there to choose from, and they can generate different types of content.
Click on the plus (+) icons below to explore some examples of what AI can generate.
Activity overview: What can genAI tools do?
An image is displayed of 5 selectable icons, as follows:
- Text content
- Images
- Sound and video
- Code and 3D models
- Research discovery
The interactivity
The user can click on the icons to reveal examples of what genAI tools can generate.
Revealed information
- Text content: AI tools that generate text responses are trained on a massive amount of data. The training process involves consuming large amounts of text from books, articles, and websites, then analysing the text to find patterns and relationships in human language. Once it is trained, it can create new text based on an analysis of human language. These tools can translate languages, summarise long passages of text, answer questions, provide data insights, write jokes and more. Text-based genAI tools can produce a range of content, such as essays, blog posts, reviews, poetry, scripts, reflective statements, news articles, role-playing scenarios and product descriptions.
- Images: Image based genAI tools can create novel images from text prompts (e.g. original images for your slide deck, or a cartoon illustrating a complex biological mechanism). You can ask the model to generate images that have never before been drawn or seen. You can also ask for specific styles of imagery (e.g. impressionist, graphic design, art deco, digital images).Not only can they produce diverse images in various styles, they can also generate images of everyday objects, landscapes, abstract art, and mashups combining multiple objects and concepts.
- Sound and video: Sound and video based genAI tools can generate content such as music compositions in a range of genres, sound effects for movies and games or human-looking avatars. AI music generators analyse music tracks and metadata (artist name, album name, genre, year song was released, etc.) to identify patterns and features. They can also be trained on lyrics associated with songs and some can produce sheet music. Other sound and video based tools can mimic voices to generate voiceovers, create video content and even edit videos including cutting clips, applying artistic styling and enhancing the resolution.
- Coding and 3D models: Coding and 3D model based genAI tools can create software applications and scripts, recommend improvements to code and even find errors and fix bugs. They can assist in designing websites and templates, creating models, characters and animations, architectural designs and interior layouts.
- Research discovery: Research discovery based genAI tools can automate parts of the research process and make long, complex texts easier to decipher by extracting key information or summarising a paper. This type of AI often analyses research papers that users upload. They can also help with literature review mapping and citation.
Remember and reflect
Key takeaway
GenAI represents a class of models that leverage machine learning techniques (learn more) to generate new and contextually relevant data in response to user generated prompts or questions. As with any technology, you’ll need to be aware of the limitations and remember that this technology is rapidly evolving.
Consider
From the above 5 types of content, which ones do you think will be most helpful for your studies?