Generative AI
On this page, you can learn more about how to use generative AI in a critical and responsible way in connection with your studies at EK, while also learning how to work effectively with chatbots.
What is generativ AI?
When we talk about AI, we are surrounded by mysterious terms and acronyms: chatbot, LLM, ML, NLP, just to name a few that you may have encountered.
Countless books and articles have been written about each of these topics, but our focus here is not so much on the underlying technology, but rather on its use in a study context. However, it does make sense to briefly introduce these concepts, as they are fundamental to understanding what it really means to work with generative AI.
LLM stands for large language model and can, in some ways, be compared to the neural networks that make up our brains. They contain millions of “neurons” that have been trained on vast amounts of data using machine learning (ML). Language models have existed for many years in various forms, but experienced a major technological breakthrough in 2017, which forms the foundation of the current wave of generative AI.
For years, we have encountered chatbots from companies and public institutions that (with limited success) have attempted to simulate human dialogue in order to save staff resources. With the simple text interfaces offered by, for example, OpenAI’s ChatGPT or Google’s Gemini, it has now finally become possible to create something resembling a real conversation with a computer.
This is made possible through natural language processing (NLP), which, in simple terms, enables computer programs and chatbots to understand and even generate text and speech. It is important to understand that the responses you receive are merely expressions of statistical probability, based on the data and training the language model has been given.
In simplified terms, chatbots can be described as probability machines that produce the most likely output given a specific input. In other words, there is no critical thinking or true understanding of the context or the world we live in, this is an important point to keep in mind when working with them.
Key Points
- ChatGPT, CoPilot and Gemini are called large language models
- They are "probability machines" that generate the most likely output based on their training.
- Language models are trained on large amounts of data
- The models do not think critically!
Remember to think critically Bias, hallucinations, personal data, and copyright
When using AI in your studies there is a list of things to be aware of. Here we will briefly introduce four of the most important things to be cautious about, bias, hallucinations, personal data and copyright.
Key Points
- Big language models are not neutral, but have bias and values built in.
- Double check always facts: Chatbots can “make up truths”
- Be cautious with what you upload to an AI and never share personal data!
Prompting How to use chatbots
The internet is flooded with guides on prompting and “how to become a prompt engineer” courses. The word “prompt” has taken the world by storm, and what was once a niche term reserved for programmers and data enthusiasts is now everywhere.
It can be debated how much sense all this focus and hype around prompts actually makes—partly because there are so many different language models trained in different ways, and whose functionalities are continuously evolving, but also because there is a certain amount of superstition and magical thinking involved, where anecdotal successes are elevated to methodological truths.
Here, we will try to keep it simple and offer just a few very brief tips on how you can get started with prompting chatbots more effectively. However, keep in mind that this should be taken with a grain of salt, and that everything is subject to constant change as the technologies continue to develop.
Key Points
- You can get better outputs if you use prompting techniques
- Prompt engineering is not an exact science
- Different prompting techniques are suited for different aims
Prompts for use in your studies
The prompts are designed based on the R-T-F framework and the chain-of-thought method. They are organized according to the topics you find here on Study Skills, and can be used with your preferred AI study partner/language model.