Getting Started with Generative AI in Research
Open course. March 2025 edition
Welcome to “Getting Started with Generative AI in Research.” This course provides a high-level understanding of technologies behind GenAI, explores use cases in research, and highlights ethical implications. While the course is primarily designed for Ph.D. students, it might be useful for more experienced researchers and the general public as well. What sets the course apart from many other offerings is that it’s built around research papers, which I hope will help learners develop a critical, evidence-based understanding of how to work effectively and ethically with generative AI. Please, see Course philosophy for more details about my approach.
Course structure and navigation
Generally, I assume that you will follow the course content in the provided order, which allows later content to build upon previously discussed material. At the same time, the course was designed with flexibility in mind and you should be able to focus on specific areas of interest while skipping or skimming over parts that may be less relevant to you.
1. What is Generative AI
This section provides a high-level understanding of the technologies behind Generative AI. While it might be tempting to directly jump to the graphical interface of ChatGPT, Claude, or Gemini, I believe that having a conceptual understanding of underlying technology would ultimately make you a better user and would also help critically assess claims about possibilities and limitations of GenAI.
2. How to talk to Generative AI
This section explores both the strengths and limitations of generative AI, along with strategies to overcome these limitations. Such strategies are often referred to as prompt engineering, though we discuss why one should be careful with this term. The principles covered in this section are broadly applicable and can be used across many different use cases beyond those specifically discussed in the course.
3. What can it do
This section focuses on research-specific applications and covers the whole cycle from idea generation to presenting your research results. We also discuss how you can use generative AI to support you in your PhD journey. The content includes practical examples and specific use cases relevant to researchers.
COMING SOON
4. Why you should be careful
Generative AI opens many exciting possibilities for researchers; however, we need to be thoughtful about potential risks, especially given that there are no established practices for using GenAI. Since technical limitations and weaknesses are covered in previous sections, here we concentrate on the ethical implications that researchers should keep in mind.
COMING SOON