Tonight at my church we had a discussion about Pope Leo’s recent encyclical on AI and human dignity. It was packed, and very meaningful to hear people from so many walks of life reflect together on this challenge we’re all facing.
Afterward, one of the elders asked if I could write down one of the ways I described the encyclical’s impact on me at the event. It was in the form of three questions. They are mainly geared toward personal experience, but they also have implications for politics and policy. These are questions, drawn from Pope Leo’s words, that I have been carrying with me in my own experiments with AI tools recently.
Is it keeping you from a relationship you could be having?
One person tonight mentioned how she was about to ask a chatbot how to get out a stain. But instead she asked her friend and her mother, and it deepened her relationships with both in the process. It ended up being a chance for passing on intergenerational knowledge and finding commonality. While chatbots do what we tell them, or pretend to try, relationships require reciprocity and accountability.
I have been recently experimenting with AI research assistants for a book project. Whenever possible, I try to have my research assistants to do collaborative work and work tied to their career trajectories, not supporting me in my writing. So here the AI collaboration feels more additive than like a replacement, and it further allows my human RAs to pursue their own interests. But I do miss my relationships with interview transcribers I used to hire, before I let AI take that over.
Is it taking on a decision that you should be making?
Are you letting an AI make important choices for you before you really get the chance to weigh the options for yourself? This concern extends for the pope to decisions about life and death in war. As one person tonight said, the closer a decision is to life and death, the farther away AIs should be.
I have sometimes caught myself tempted to let AI do my writing for me. But while the little decisions about a word here and there are probably not terribly important in the scheme of things, they are my profession and my craft, and I have to stand by them. For other people, I realize, writing may not involve their most important decisions, and AI assistance could free them to deliberate more fully where they need to.
Is it keeping you from an opportunity to learn that would help you grow?
One person said he had just left his job as a software engineer because of the pressure to automate work that he enjoyed learning to do himself. Another, a teen, mentioned using it to not have to think about proportions in a recipe. (The younger folks were strikingly more AI skeptical than the older ones.)
I have been wrestling with how much to use AI tools to augment my research methods. Can I let an agent write the code for a commonly used kind of data analysis—code I don’t yet know how to write myself? What do I lose by not going through the process of learning it? Is the tool helping me learn, or preventing me from doing so?
I hope these questions are helpful.