I got a message recently on LinkedIn from someone who sent me a GitHub repo of futures-related SKILLS.md files put together.
This got me wondering: why would one want to give over the actual process of figuring things out to a set of agentic SKILLS.md files so that, presumably, an LLM could like..do
the thing that one (human) would normally have to think through oneself.
It seems a deprivation of the most interesting work one could do if one were oriented towards imagining futures.
The fun part of the work is doing it and feeling the poke of something unexpected or curious or just odd enough to take the conversation down a path that was never anticipated. Or finding something unexpected in the process, such that one feels a kind of “becoming” : a sense of evolving understanding and insight, or just “the new” arriving on its own.
When that happens, that feels like Imagination at work.
The other day Howard sent me a link to an interview with
Mike Cannon-Brookes who offers one of the more even-keeled takes on the role of AI and in so doing he introduces a term he heard: “Imagination asymmetry” — he suggests that the ability to create and think may become the new great advantage — something different from compute or inference or token maxxing.
Which makes me think: if imagination becomes more valuable, what happens when we delegate to an LLM or stash of SKILLS.md files the activities through which we exercise and develop it? Losing hold of one's ability to discover feels existential to me. That discovery happens while you’re doing the work, rather than having delegated it
elsewhere. You start with the thing you're wondering about, a question, the contours of such as a brief. Then you allow yourself, on a kind of journey, to encounter something peculiar, then follow it and you just may find yourself thinking about something you couldn’t have asked for when you set off on the journey. If you delegate the exploration, you just may remove those collision-occasions. You receive the agent’s account of what it found, with its selection of what deserved attention.
Maybe what we've been calling “futures work” doesn't quite situate what it is that we've been trying to do. The work isn’t just applying methods to imagine what comes next. It’s allowing oneself to leave the old language and ideas behind while venturing into somewhere unexplored, having the time to allow for encounters in
this new terrain, learning to notice what matters — and helping other people notice it, too.