This post introduces a trial blog series called “Practically Prompted” – an experiment in using large language models (LLMs) to write a Practical Ethics blog-style post, with some light human commentary about the output.
So, why try this? The experiment is driven by several key motivations:
- To Test a New Tool: We want to see if LLMs can do more than just mimic the style of our blog. Can they actually illuminate the moral tensions in our world and provoke new ways of thinking?
- To Keep Pace: The news cycle and technological change are relentless. LLMs could offer a way for academics to respond in real-time and help the public think clearly about pressing ethical issues.
- To Invite Reflection: By publishing AI-written posts alongside human commentary, we can track both the promise and the pitfalls of using these models. This allows us to explore questions like:
- What makes a practical ethics post “thoughtful”?
- What moral frameworks do LLMs favor and which do they neglect?
- Where do they overreach, equivocate, or reinforce unexamined assumptions?
- Are readers even interested in ethical analyses by AIs?
- To Stimulate Discussion: This series raises important questions about expertise, authorship, and human judgment. It prompts us to consider the future of collaboration between human philosophers and AI, including what might be fruitful and what could be corrosive.
A final note on the method for this trial: The LLM-generated posts will be the direct, untouched output from ‘o3’, currently OpenAi’s most advanced and research-oriented model. A more involved prompting process would certainly improve the result, but for this experiment, the goal is to see what the model can do with a simple set of instructions. This allows for consistently tracking the model’s capabilities for this kind of task over time.
Click here to read the first post in the series.