In many ways it's good for my work that LLMs can't write as well as they can code. I can't afford to ship a mid line of copy when my business is a direct reflection of my self, my output more exec comms than consumer copy. So I'm learning to turn to it for systematizing new ideas and ways of working; I know I can be better allocating my unpaid time (too much on admin & ops, not enough on sales & R&D). But to integrate its ideas into my existing systems is still Work because let's be real, it usually offers too much. So my ideas-to-systems process looks like: talk to Victoria about building pipeline, go to sleep, wake up and whatever's still bouncing around in the morning I'll continue, document it in the codex, have it on deck for when Victoria 2.0 update rolls around. That's a lot of lag time, and my business is just one person! Imagine the enterprise scenario... actually, yes, I just read something on that: [Frank Chimero on bypassing the mess](https://frankchimero.com/blog/2025/beyond-the-machine/): > "Faced with the story of AI labor displacement, our first instinct as technology workers wasn’t to protect one another, but to search for ways to use the tools to replace our collaborators. > > The fractures fell neatly along disciplines: engineers using AI to wish away designers, designers wishing away engineers, product managers wishing away both. In this climate, AI becomes frenemy identification technology, another way to avoid working together. It’s always easier to grab a tool and bypass the mess of coordination, even if that means doing more and doing it alone. AI lowers the barrier to working outside your lane, and sure, that could mean more overlap between disciplines, but right now, we’re mostly boxing each other out or stepping on one another’s toes." I'm trying out the inverse this fall, working with J. and M. on building a side project involving agriculture, local businesses, and policy (soooooooo excited). Seeing what happens when you condense product-ops-design-eng-brand-content into three people, squish them in a box of a Midtown cafe with a problem to work on. The 2.0 of my practice to [[Host bullpens to surface new ideas and refine thinking]], taken to the depth I always aimed for — commercial & product orgs co-creating rather than just trading notes. I call it the first makers cohort of The Fermentation Room, by Skin Contact Studio®. (bc Victoria says *everything has a place in the system*) A test in the speed of human adaptation, strip the red tape. [Sarah Constantin writes in "The Great Data Integration Schlep"](https://sarahconstantin.substack.com/p/the-great-data-integration-schlep): >If you’re imagining an “AI R&D researcher” inventing lots of new technologies, for instance, that means integrating it into _corporate_ R&D, which primarily means big manufacturing firms with heavy investment into science/engineering innovation (semiconductors, pharmaceuticals, medical devices and scientific instruments, petrochemicals, automotive, aerospace, etc). You’d need to get enough access to private R&D data to train the AI, and build enough credibility through pilot programs to gradually convince companies to give the AI free rein, and you’d need to start virtually from scratch with each new client. This takes time, trial-and-error, gradual demonstration of capabilities, and lots and lots of high-paid labor, and it is _barely being done yet at all_. > > I’m not saying “AI is overrated”, at all — all of this work _can_ be done and ultimately can be extremely high ROI. But it moves at the speed of human adaptation.