Let me step into that historical moment of being 2nd who give it a like, and first to comment. What you just said in this post is so true to me before AI era that I'm very happy for you and kipping fingers crossed for you to succeed! We are on the verge of one more cognitive challenge in the history of humanity, like previously it was reading, then it was computers, now is AI. AI pushes the boundaries of humanity cognition further than they currently are so what you are up to is interesting from the human point of view. BTW thought challenging company name 🙌
Thanks, Tomasz — and congratulations on the historic first-comment moment 🙂
You’ve touched on something I keep coming back to: these four kinds of work existed long before AI. What changed is not the structure of a company, but which parts of it a system can carry.
We could automate parts of execution before, but people still had to interpret context, handle exceptions, and connect one process to another. AI can now begin to carry that layer too, and that changes how much glue and execution can move away from people.
Your analogy with reading and computers fits well, because each of those shifts moved the boundary of what people had to do themselves. This one begins to move contextual interpretation into the system — the layer that kept people in the middle of almost every process.
ZHC is my attempt to see where that lands in practice, not just on a map.
It resonates with me a lot. I run a 50+ person company. A year ago, it was 100. I’m so tired of being the glue and the executor. Sometimes it feels as if there is no time left for judgment, which is the part only I should be doing.
AI finally lets me move some of that work off people’s desks. Liberating. But transforming operations at this scale is hard. The glue is everywhere, and every process is attached to people, habits, and other processes.
Thank you, Katherine. Your point about having no time left for judgment captures it perfectly. The person who should be deciding ends up carrying context, pushing the work along, and holding the whole thing together. That is the glue, and it lands on whoever is connected to everything, usually the founder.
You are rebuilding a living company of 50+ people, where the glue has piled up over years and grown into people, habits, and systems. ZHC comes at it from the other side, building a new company so a human never becomes that glue in the first place. In that sense I have it easier. Yours is harder, and far more common.
The more I look at it, the more it comes down to one thing: who remembers where everything is. In most companies, people do. They hold the state of things in their heads, carry context between tools, and piece the picture together. That is the glue. As long as that memory sits in people, the founder stays the one holding it all. The glue only goes away when the state of things lives in the system itself, complete enough that the next step can run without a human rebuilding the context.
I would genuinely like to hear which kind of glue has been hardest to remove at your scale. That is exactly the ground truth this experiment needs.
Your summary is precise :). At my scale, of course, a lot of things live in people's heads, and the biggest piece is in mine. When you have a company of 50 to 100, a lot of processes are custom. You don't need heavy bureaucracy because it slows you down, but when there is no real workflow, there is nothing to hand to an AI workflow either.
On top of that, some people are resistant to change. With AI they are even more scared, because they feel they could lose their jobs.
Between us, sometimes I think it would be easier to start from scratch than to transform what we have.
Coming back to this a bit late, but your last point stayed with me.
It feels a lot like renovating a house. Once you start opening things up, you find problems and dependencies nobody expected, and at some point you realize building from scratch might have been easier, sometimes even cheaper.
I suspect AI transformation in an existing company works the same way. The glue is hidden in people, habits and processes, and you only discover most of it when you try to remove it.
Your comment kept coming back to me while I was writing, so it ended up shaping two follow-up pieces on glue — one out tomorrow and one next week.
Let me step into that historical moment of being 2nd who give it a like, and first to comment. What you just said in this post is so true to me before AI era that I'm very happy for you and kipping fingers crossed for you to succeed! We are on the verge of one more cognitive challenge in the history of humanity, like previously it was reading, then it was computers, now is AI. AI pushes the boundaries of humanity cognition further than they currently are so what you are up to is interesting from the human point of view. BTW thought challenging company name 🙌
Thanks, Tomasz — and congratulations on the historic first-comment moment 🙂
You’ve touched on something I keep coming back to: these four kinds of work existed long before AI. What changed is not the structure of a company, but which parts of it a system can carry.
We could automate parts of execution before, but people still had to interpret context, handle exceptions, and connect one process to another. AI can now begin to carry that layer too, and that changes how much glue and execution can move away from people.
Your analogy with reading and computers fits well, because each of those shifts moved the boundary of what people had to do themselves. This one begins to move contextual interpretation into the system — the layer that kept people in the middle of almost every process.
ZHC is my attempt to see where that lands in practice, not just on a map.
Really glad you’re following along.
It resonates with me a lot. I run a 50+ person company. A year ago, it was 100. I’m so tired of being the glue and the executor. Sometimes it feels as if there is no time left for judgment, which is the part only I should be doing.
AI finally lets me move some of that work off people’s desks. Liberating. But transforming operations at this scale is hard. The glue is everywhere, and every process is attached to people, habits, and other processes.
Thank you, Katherine. Your point about having no time left for judgment captures it perfectly. The person who should be deciding ends up carrying context, pushing the work along, and holding the whole thing together. That is the glue, and it lands on whoever is connected to everything, usually the founder.
You are rebuilding a living company of 50+ people, where the glue has piled up over years and grown into people, habits, and systems. ZHC comes at it from the other side, building a new company so a human never becomes that glue in the first place. In that sense I have it easier. Yours is harder, and far more common.
The more I look at it, the more it comes down to one thing: who remembers where everything is. In most companies, people do. They hold the state of things in their heads, carry context between tools, and piece the picture together. That is the glue. As long as that memory sits in people, the founder stays the one holding it all. The glue only goes away when the state of things lives in the system itself, complete enough that the next step can run without a human rebuilding the context.
I would genuinely like to hear which kind of glue has been hardest to remove at your scale. That is exactly the ground truth this experiment needs.
Your summary is precise :). At my scale, of course, a lot of things live in people's heads, and the biggest piece is in mine. When you have a company of 50 to 100, a lot of processes are custom. You don't need heavy bureaucracy because it slows you down, but when there is no real workflow, there is nothing to hand to an AI workflow either.
On top of that, some people are resistant to change. With AI they are even more scared, because they feel they could lose their jobs.
Between us, sometimes I think it would be easier to start from scratch than to transform what we have.
Coming back to this a bit late, but your last point stayed with me.
It feels a lot like renovating a house. Once you start opening things up, you find problems and dependencies nobody expected, and at some point you realize building from scratch might have been easier, sometimes even cheaper.
I suspect AI transformation in an existing company works the same way. The glue is hidden in people, habits and processes, and you only discover most of it when you try to remove it.
Your comment kept coming back to me while I was writing, so it ended up shaping two follow-up pieces on glue — one out tomorrow and one next week.