Types of generative AI

Artificial Intelligence is the ability of computers or computer-controlled robots to perform tasks commonly associated with intelligence. Basically, if a human did it, it would be classified as intelligent. Think playing chess or analysis, spotting patterns and making predictions based on huge amounts of data. 

GenAI refers to a type of artificial intelligence that has been trained using large amounts of data to perform, mimic or imitate tasks. This technology uses machine learning to reproduce human behaviour and create new content based on the data it was trained on.

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 produce essays, blogs, scripts, news articles, reflective statements and even poetry. 

Some examples of GenAI that can create text content include: ChatGPT and Perplexity AI. Below is examples of prompts and outputs that were given to ChatGPT.

Image generative AI works because researchers and developers feed machine learning programs hundreds of millions of images and descriptions that have been scraped from the internet. The model is then able to learn and identify relationships between the images and words. These models inherit the bias of their training datasets. For example, the prompt 'a doctor' using Canva's text to image resulted in four pictures of older white males in white lab coats. 

These tools can produce diverse images in various styles, everything from anime, post-impressionism and watercolours; however, changing or tweaking a produced image can be very difficult.  

Some examples of generative AI that can create imagery include Dall.E 2, Midjourney and Stable Diffusion. They have also been worked into programs such as Adobe and Canva. 

Creating video content is time-consuming. You need a script or lyrics, recordings and then you need to edit. 

AI music generators analyse music tracks and metadata (artist name, album name, genre, year song was released, ect) to identify patterns and features. They can also be trained on lyrics associated with songs and some can produce sheet music. 

Some examples of generative AI that can create audio content include AIVA, Soundful and Murf.ai.
Creating a video typically requires the use of audio, visual and text elements. There are generative AI video programs that have been trained on some or all of these elements allowing you to control what is created. 

Some examples of generative AI that can create videos include Gen-1 Runway and Invideo.

Learning to code is similar to learning a language, and can be just as hard. Instead of searching the internet for help when coders get stuck, generative AI models can now be used to help with generating and improving code or even finding errors. 

This means that you could use a tool like GitHub Actions to reformat new code to match old, instead of spending hours manually formatting it yourself; or even build a complete website with no coding experience. But beware, it'll likely come at a cost. 

Some examples of generative AI that can create code include ChatGPT, CodeT5 and Tabnine.

There are many generative AI tools that claim to automate parts of the research process and make long, complex texts easier to decipher. This type of AI often analyses research papers that users upload to extract key information or summarise a paper. they can also help with literature review mapping and citation. 

Some examples include: Elicit, Raxter, Research Rabbit and Scite

Key Considerations

Generative AI is limited by the amount, quality and context of the data it's trained on. There is a lot of controversy over what datasets have been used to train GenAI and whether this was a breach of copyright. GenAI has created a crisis for copyright law and there are a number of lawsuits currently underway. 

Many of these generative AI tools cost money to use or to access premium features. In some cases, you can create a basic account for free or explore the tool with a short-term trial.

Many library database agreements contain clauses around the use of their content and AI, so please read the agreement first.