Braiding AI

I guess I’m carving out a space focused on AI metaphors. First it was working with Anna Mills to suggest “mirage” rather than “hallucination” as a metaphor for when AI gets it wrong. Now I’m fixated on this thing we keep hearing about the importance of keeping humans “in the loop” with AI. We need to stop this kind of talk.

Our idea of properly deployed AI should not be just that there are humans “in the loop” of AI, presumably monitoring and adjusting AI activities to align them with human priorities and values. Instead, we should be working toward a world where AI only operates when it is subordinated within a loop that begins and ends with humans. Let’s imagine a future where AI is in our loops — when and where we want it — rather than us humans being mere facilitating components in AI’s loops.

But is a loop even the right metaphor at all? All credit to Karen Bellnier for suggesting the better metaphor of a braid rather than a loop in a recent Future Trends Forum conversation. Loops conjure up ideas of logic and programming, or of transportation routes or rollercoasters, or maybe of sewing or hiking if we look to more human pursuits. But braids. Braids are better: Starting on the human head from our own hair, spiraling out to maybe include other media in creating diverse forms — from simple pigtails to the intricate patterns of complex hairstyles — but ending where we decide, always attached to us and headed where we are going. Or even bread — like the braided challah that gives us life and marks our rituals. Let us braid AI into our needs, into our goals, and into our dreams rather than resigning ourselves to just function “in the loop” of AI.

My continuing emphasis on how we talk about technologies like AI keeps butting into other posters who have narrowed their focus to how we might use AI for this or that specific purpose. Sure, they may be figuring out some beneficial uses for AI. But are their increasingly AI-centric posts really helping us grapple with how AI fits into our larger human project? Or are they just helping us accept AI as it is, without even helping us find the words we need to think about AI differently? We are just starting on this journey. We don’t need endless “how to drive” posts when we have still not yet figured out where we will be going with these new shiny tools.

I contend that how we talk about technologies like AI can have a big impact on how such technologies affect our world. Think back, for example, to our history with the automobile. While we were starting to name our cars, hum along to Chuck Berry and Beach Boys songs, and wrap our identities around our rides, automobiles took on an increasingly transformative status in our world, eventually dramatically reshaping our built environment, our natural ecosystems, and our geopolitics for the worse, all of which outcomes we had difficulty foreseeing, but which we are suffering now and seem incapable of fixing. Just like we should have with cars, we need to take care how we incorporate technologies like AI into our common discourse, and into our imagination and desires. Otherwise — as we did with cars — we risk building ourselves into a culture mostly shaped by the immediate goals of industries that benefit from technology adoption and use, but are not motivated to help us shape technology for our long-term benefit.