Generative AI Tools
ChatGPT
ChatGPT has grabbed most of the headlines since its launch in November 2022. It was created by a company called OpenAI, which started as a not-for-profit research organisation (hence the name) but is now a fully commercial company with very significant investment from Microsoft.
ChatGPT is based on a machine learning approach called 'transformers', first proposed in 2017, and is pre-trained on large chunks of the internet, which gives it the ability to generate text in response to user prompts, hence the name ‘Generative Pre-trained Transformer’.
ChatGPT essentially works by predicting the next word given a sequence of words. This is important to understand, as it is not in any sense understanding your question, and then searching for a result, and has no concept of whether the text it is producing is correct. This can make it prone to producing plausible untruths or, as they are often known, hallucinations.
ChatGPT – the newest model (May 2024)
Since launching in November 2022, there have been a number of different models, ChatGPT 4o is OpenAI’s new flagship model that can reason across audio, vision, and text in real time. Some of this functionality is not yet available to users in Ireland but is expected to be rolled out in coming months.
ChatGPT 4o, is available as a ‘Free tier’ model. This means that all registered users have free access to the most up to date ChatGPT model (with some limitations) on a desktop/laptop computer. The Free tier option ChatGPT 4o is not yet available on mobile devices. You will still need to create an OpenAI account to access Free tier ChatGPT 4o.
ChatGPT 4o currently has a context window of 128k which, in terms of word count, translates to approximately 85,000 to 96,000 words.
As it stands today, the most up to date Free tier version ChatGPT 4o has a training data cut-off date of October 2023. This means that the model’s knowledge is based on information available up to that time, and it does not include any events, developments, or discoveries that occurred after October 2023. Understanding this cut-off date is necessary to set appropriate expectations for the information provided by ChatGPT 4o and highlights the need for corroborating information when dealing with recent or evolving topics.
There is also a premium version ChatGPT Plus which costs approximately $20 +VAT per month.
Users on the Free tier will be defaulted to ChatGPT 4o on sign-in, with a limit on the number of messages they can send using ChatGPT 4o which, according to OpenAI, will vary based on current usage and demand. When unavailable, Free tier users will be switched back to ChatGPT-3.5.
Users are notified once they have reached the limit and will be invited to continue the conversation using ChatGPT 3.5 or to upgrade to ChatGPT Plus.
The following tools have shared usage rate limits that are separate from the ChatGPT 4o text rate limit:
- Data analysis with ChatGPT
- File and image uploads
- Browsing with ChatGPT
- Discovering and using GPTs
- Vision
If you reach a rate limit for any of these tools, you will need to wait until the system allows you to begin using the tools again.
Users paying for ChatGPT Plus have access to a version that allows them to build their own GPTs , generate images using DALL-E, as well as having access to higher rate limits in a chat.
As of May 13th 2024, ChatGPT Plus users will be able to send up to 80 messages every 3 hours on GPT 4o and up to 40 messages every 3 hours on ChatGPT 4. Please note that unused messages do not accumulate (i.e. if you wait 6 hours, you will not have 80 messages available to use for the next 3 hours on ChatGPT 4).
Introducing OpenAI o1-preview – available 12th September 2024 to premium customers only
Please note that o1-preview and o1-mini are preview models, and their availability described at links below may be subject to changes.
Learn more about the latest models OpenAI o1 and Open AI o1-mini
What conversation topics are appropriate for OpenAI’s o1 models?
OpenAI o1-preview and o1-mini models do not have access to the following advanced tools and features:
- Memory
- Custom instructions
- Data analysis
- File uploads
- Web browsing
- Discovering and using GPTs
- Vision
- Voice
You will need to switch over to GPT-4o to access these tools. (Source 17/09/2024: Using OpenAI o1 models and GPT-4o models on ChatGPT)
Attribution: A Generative AI Primer JISC National Centre for AI Version 1.3 – 2nd Jan 2024. License: CC BY-NC-SA
Page updated: 18/09/2024
A type of neural network architecture used in deep learning models, particularly effective for understanding the context in language tasks.
The context window in ChatGPT refers to the amount of text (in tokens) that the model can consider when generating responses. A token is a unit of text, which can be as short as one character or as long as one word. The context window determines how much of the conversation history, including the current input and previous interactions, the model can "remember" and use to generate coherent and relevant responses. GPT-4o currently has a context window of 128k. According to OpenAI, in terms of word count, this translates to approximately 85,000 to 96,000 words. Understanding the context window is crucial for effectively managing long interactions and ensuring that important details are preserved for as long as needed within the conversation.
