Research Drops

Conversation continues to circle the tension between AI versus human output. When should AI be allowed and to what extent? How should we signal AI's involvement and what level of detail ensures meaningful attribution? We can and should continue these important conversations, whilst being mindful of the fact that when, where and how we use AI is not only dependent on the nature of the task itself, but on our ability to utilise the technology meaningfully, accurately and honestly.

Ethan Mollick amongst others, has written on his blog One Useful Thing about the scissor effect (a term originally coined Scott Alexander). It describes a growing divide not just amongst those who use AI versus those who don't, but between those who use AI to extend their capabilities and those who use it as a crutch without learning anything. Cognitive offloading is something I think about a lot these days, for my children and for my own work.

This Research Drop experiment is an attempt to work through some of those concerns personally and professionally. The question I've been circling is this: what might be gained if we step outside of the binary conversation poles of 'human vs AI' production and instead focus on how AI can extend our human capability. Where does it make sense for us and our work, and where does it not? If we share our approaches, our methodology and our learnings not just in the name of transparency (although that matters too), but in the pursuit of better understanding of where AI genuinely extends our thinking and where it quietly replaces it, then perhaps we can move beyond policing the boundary and start learning from what happens either side of it.

This experiment is one small, imperfect attempt to do that in the open.