[Image taken from Pixabay]Artificial Intelligence (AI) is a field of computer science that focuses on creating intelligent machines. These machines can mimic human thinking and learning processes, allowing them to solve intricate problems, make decisions, and perform tasks autonomously.
[Open AI, 2023]
It is still early in the development of Generative AI and so this guide may quickly become outdated. This guide highlights a variety of tools that you may wish to use to support you in your studies. The University of East London does not support or endorse any of the AI tools mentioned in this guide.
Always refer to your Tutor for the latest guidance on using AI.
Generative AI refers to machine learning models that are trained on existing data to create original output such as text, videos, images and data in response to prompts.
Perhaps the most well known Generative AI tool is ChatGPT, developed by OpenAI (GPT stands for Generative Pre-trained Transformer). ChatGPT allows users to engage in dialogue to seek information, ask follow-up questions and, crucially, challenge incorrect premises.
OpenAI currently has two versions of ChatGPT available to the public:
GPT-3.5 - Released in March 2022 and currently freely available.
GPT-4 - Released in March 2023 and is only accessible via subscription.
Both versions lack knowledge of events that have occurred after September 2021, when the vast majority of its data cuts off (OpenAI, 2023).
ChatGPT-5 is currently in development (as of September 2023).
Although the ChatGPT interface is the most well-known, there are a variety of other Generative AI tools available to use.
When asking ChatGPT what its limitations are, it lists the following:
1. Lack of real-time information: My knowledge is based on text data up to September 2021, which means I'm not aware of events, developments, or information that has occurred or become available after that date.
2. Inability to access the internet: I cannot browse the internet or access external databases, websites, or real-time information. My responses are generated solely from the knowledge I was trained on.
3. Limited context: I generate responses based on the input I receive, but I do not have memory of past interactions within a single conversation. This means I may not remember details or information from earlier in the conversation.
4. Potentially biased information: My responses are generated based on the data I was trained on, which may include biases present in the training data. It's important to critically evaluate and cross-reference the information I provide.
5. Lack of personal experience or consciousness: I am not a sentient being and do not have personal experiences, emotions, or consciousness. I do not have opinions, beliefs, or desires.
6. Inappropriate or harmful content: I am programmed to avoid engaging in discussions that involve illegal activities, harm to others, hate speech, or explicit content. However, I may not always detect and prevent such content effectively.
7. Limited creativity and imagination: While I can assist with creative tasks to some extent, my creativity and imagination are limited compared to humans. My responses are based on patterns in the text data I was trained on.
8. Inaccuracies and errors: I can provide information to the best of my knowledge, but I am not infallible, and there may be inaccuracies or errors in my responses. It's important to verify information from multiple sources when making important decisions.
9. Difficulty with some tasks: I may struggle with tasks that require a deep understanding of context, nuanced emotions, or complex decision-making, as my responses are primarily based on patterns in text data.
10. Repetition and verbosity: I may sometimes provide long or repetitive responses, and I may not always get straight to the point. Users may need to clarify their questions or request more concise answers.
11. Limited language proficiency: While I can communicate in multiple languages, my proficiency may vary, and I may be less accurate or fluent in languages other than English.
Overall, while I can be a helpful tool for information retrieval and generating text-based responses, it's important to recognize my limitations and use critical thinking when interpreting the information I provide. [OpenAI, 2023]
You must follow the latest guidance on using Generative AI as provided by your School.
If you do use Generative AI for your assessed work, you must acknowledge that you have used it (see Referencing). You should ensure you verify all information provided by Generative AI and not just assume it to be accurate.
See our page on Using AI Critically for key considerations when using Generative AI.
Do be wary that your personal data may be vulnerable if used in interacting with Generative AI.
The AI landscape is constantly evolving and guidance may quickly become outdated.
OpenAI. (2023). ChatGPT 3-5 [Large language model]. https://chat.openai.com/