Atrophy of effort

---

I’ve been asking myself how much of my thinking I’ve been willing to surrender. With the conveniences of LLMs seeping into pretty much every part of knowledge work, I’ve been wondering what the cognitive costs might be.

What if all the AI tools we use asked us to frame a manifesto or a constitution of sorts, for ourselves, before we’re given access to them?

So if I’m using an AI search engine for mainly research, I’d be careful to not ask for or receive any ideation or brainstorming suggestions, even if I wanted to cut down on time for these tasks. And if at all I do need some help, I’d be asked to amend my constitution.

Something like this isn’t new though. Most of our popular LLMs allow us to provide instructions and customise our experience with them, but hidden under the hood, in the Settings menu, where it’s more of an ‘nice-to-have’.

I’ve wished for this kind of UX pattern because of the trends in AI adoption I’m seeing, particularly as a designer and a recent student at that. A gesture like this that focuses on intentional, frictional use could help creatives frame their own guardrails.