We know that something is missing by the shape of the negative space around our new tools. Anyone who’s gotten deep in use with them feels simultaneously radically enhanced and disoriented and diminished.
People have forgotten (or never learned) the techniques for how to succeed. The literature tells us how to succeed, how to set goals and build incrementally towards them. Our biology has been tuned over millennia to reflect on these goals, to feel satisfied not only in their completion, but in the process of having completed them.
A farmer who tills a field looks proudly upon his accomplishment even before any fruit has been borne. A carpenter delights in the construction of a home not only for its utility or the particularly clever techniques in execution of the craft (e.g., a wooden joint or a meeting between the foundation’s masonry and the wood of the structure), but in the transformation of his body through that process, through his participation in the final creation. A hunter delights in the hunt even without a kill.
AI has changed the process of working dramatically by yanking us from the continual feedback loop of development of ourselves through the outside world, the coupled processes of self- and external development. We can tell because although everyone who has benefitted from AI loves the power that comes from working with new tools, no one loves the feeling of it.
The tools don’t respect human cognitive bounds and processes. They’re not well-matched to human memory and the way that the body works. It’s impossible to maintain “flow” (see Csikszentmihalyi) within these systems. Human beings feel a combination of emotions: existential dread (what’s my purpose anymore?); guilt (why aren’t I succeeding more?); and resentment (I’m more productive, so why do my superiors (and myself) feel like I’m not doing enough?). We are getting a lot done and are tired for it, but the fatigue isn’t the same; rather than merely wrought, we feel extracted from, depleted.
There’s a claim circulating that we’re moving from a literary form back to orality, but that’s wrong: we are moving to a new thing that hasn’t been defined yet. AI has given us today what writing gave the cultures it overtook: a powerful new conveyance whose cognitive consequences we can’t unexperience. As then, the wisest people (Plato among them) paused at literature’s adoption as the primary mode of thought and discourse. Today, we should pause to reflect on the effect that the tools have on us.
Medieval English kings demanded that all young men participate every Sunday for two or more hours in developing the skills of the longbow. That technology was totalizing and laid the foundation for English regional supremacy for centuries, but it transformed not only the entire economy around it, but also the very bodies of the citizens: for eternity, when skeletons of these people are found, the signature of the asymmetrical development of the musculature and its effects on the spines of these men is visible.
We are imposing on ourselves a similar transformation today, although perhaps less immediately visible in our physical presentation. Most critically, these transformations are happening not only to our bodies, but to our minds, the very instruments available to us to reflect on the transformation. We’re also incentivized against reflecting on these changes because, one, we might not like what we see, and two, our livelihood increasingly depends on adoption of these disruptive tools. To paraphrase Sinclair, a man will not understand what his salary depends on his not understanding.
People are doing things that don’t make any sense. We are working with dozens of agents concurrently — the scent of productivity, of impact, at the tips of our senses — and then deleting the logs (Walsh & Ungson taught us better than to do this). We are starting projects and threads within them without finishing any. We make sketches of ideas, an intellectually satisfying exercise, but we struggle to close the loop, to bring any to fruition. We’re prototyping like never before, realizing ideas we’d have only dreamed of starting, yet aren’t feeling the joy of completion that our farmer, carpenter, and hunter ancestors did.
But it’s not any one of our faults: the new tools have two distinct yet compounding characteristics. The first is that they don’t have the affordances (see Don Norman) that suggest we close these loops. The second is that the tools aren’t yet (and may not ever be) good enough to let us finish what we start. On that second point, it’s worth considering whether, even if they did, we would feel the same satisfaction in completed works. I’m skeptical unless we develop a scaffolding or a substrate that allows us to see, to feel our unique, human contribution to the problems we’re solving and the solutions we guide for them, so that we can feel in ourselves an internal progress as well.