policy_ai.qmd

Artificial intelligence (AI) tools are increasingly well-known and widely discussed in the context of data science. AI products can increase the efficiency of code writing and are becoming a common part of the data science landscape. For the purposes of this course, we **strongly recommend that you _do not_ use AI tools**.  There is an under-discussed ethical consideration to the use and training of these tools in addition to their known practical limitations. However, the main reason we suggest you not use them for this class though is that leaning too heavily upon AI tools is likely to negatively impact your learning and skill acquisition.

You may have prior experience with some of the quantitative skills this course aims to teach but others are likely new to you. During the first steps of learning any new skill, it can be really helpful to struggle a bit in solving problems. Your efforts now will help refine your troubleshooting skills and will likely make it easier to remember how you solved a given problem the next time it arises. Over-use of AI tools can short circuit this pathway to mastery. Once you have become a proficient coder, you will be better able to identify and avoid any distortions or assumptions introduced by relying on AI.

**AI Resources**

- Pratim Ray, P. [ChatGPT: A comprehensive review on background, applications, key challenges, bias, ethics, limitations and future scope](https://www.sciencedirect.com/science/article/pii/S266734522300024X). **2023**. _Internet of Things and Cyber-Physical Systems_
- Trust, T. [ChatGPT & Education Slide Deck](https://figshare.edgehill.ac.uk/articles/presentation/ChatGPT_Education_Slide_Deck/21901629/1). **2023**. _National Teaching Repository_
- Csik, S. [Teach Me How to Google](https://ucsb-meds.github.io/teach-me-how-to-google/#1). University of California, Santa Barbara (UCSB) Master of Environmental Data Science (MEDS) Program.