Supervisor meeting – what was discussed, what feedback and ideas did you receive? How are you moving forward with this feedback and what are the next steps?
In my first supervisor meeting, James Garforth and I discussed our backgrounds, research interests, and different possibilities for my project.
I had been very invested in doing my AI rights boardgames, but upon internal reflection and in starting to brainstorm the idea, I realised that the creation of a board game with little game design experience could prove very challenging. But perhaps more poignant, was that James’s question of ‘what do you want to do after this?’, brought into focus the idea of doing a project aligned with my desired career. Moreover, I wanted to marry my two major research interests – information systems design, and AI ethics.
I floated my second idea past him – to look at traditional software development methodologies and ‘alter’ them to include considerations of ethics in their workflows, possibly making it easier to conceptualise ethical consideration within SW development projects, in ways which would be familiar to developers. This would also be well within my scope of skills, given that I had conducted literature reviews before – I consider this one of my better hones skills. Given his informatics background, he found this idea interesting, and began suggesting different ways I could conceptualise the idea. For example, James suggested that I focus on a specific domain of software development methodologies, that caters towards NLP, and then look at two different methodologies and compare them, integrating ethical considerations throughout / considering how easily ethical considerations could be incorporated.
I left the meeting brimming with this idea. In taking the next steps, I started to create a new miro board considering how this kind of literature review could be structured. These are the topics which emerged from my initial sweep of the literature:
- Historical papers on the relationship between AI & SW engineering
- NLP software development compared to traditional software development methodologies, user stories & testing phases
- Human-centred AI design, and different frameworks that have been developed to engender ethical considerations in software development
These topics gave me a starting point for thinking about what the structure would look like, but also left me with some questions which I’d like to address in my next supervisor session:
- Should this paper take the format of a systematic lit review?
- Should I include other factors that may affect software development, such as organisational factors etc?
- Should I incorporate software project management frameworks as well as development methodologies?
- How should I (or should I at all) address the body of literature around using AI to improve software development methodologies?
- Should I suggest my own model at the end, or just piece together different models, or not suggest a model at all (rather gathering insights)?
Aside from getting these questions answered, I also a want to take a deeper dive into the literature, and continue considering what kind of structure the paper could take. While I loved my initial idea and will continue to hold it as a consideration, I am also invested in practical and transactional data and AI ethics – I believe this idea will serve me in broadening my research interests and scope, and that I can make a meaningful contribution.
Snippet of my Miro Board conceptualising this idea – unfinished but in progress (cant upload higher resolution bc of restrictions, but can be accessed here: https://miro.com/app/board/uXjVNlEmelo=/?share_link_id=106394061543 ) 🙂
