Mark Spall

Almost everyone misses what is being built at The Grove and I don't blame them - it's a big ask to see out of the boxes of current society. So here's a big picture slide: Sovereign people, Sovereign organisations, Sovereign society. It's possible now. Go and find out more at max-unleashed.com/grove/buil… - and I hope the songs are inspiring and open a little something in you.

How do you split an organisation in two without a fight over the client list? Or join two without one swallowing the other? Most mergers fail. Splits are fought over. Both usually go wrong for the same reasons: decided out of sight, done in one go, and finished before everything has somewhere to land. Living cells do the opposite, and they rarely get it wrong. A cell copies its whole genome before it divides. It holds the two copies side by side inside one membrane. And it won't separate until every last chromosome is attached to the side it's going to. I've written a new chapter of The Grove's book on how whole organisations divide, bud off a part, join together and close, borrowing the order of things from biology. One rule sits under all of it: nothing separates or joins until everything it touches has somewhere to land. It has to be built before we go live, so this is the moment to get it right. I'd value your eyes on it, especially if you've lived through a merger, a split or a spin-out. Where would this break? What have we missed? markaspall.github.io/the-gro…

Hey - I've been reading a lot of the material on p2p foundation. It's really valuable. Thanks for all your efforts behind it. I thought I'd use it as a measure for The Grove. I'm after peer review, so I attempted to do it myself! markaspall.github.io/the-gro… @mbauwens

The Squeeze on the Mind Something I keep noticing I have a hunch I can't shake, and I don't yet know whether it's true. It comes from watching social media, not from any study, so treat it as an observation rather than a finding. The feeds feel increasingly written by no one. Posts arrive polished, balanced and oddly familiar. Replies agree in the same cadence. More and more of what passes for thought online seems to have been produced rather than thought. That in itself isn't new or interesting. What interests me is what it might be doing to the people behind the posts. A squeeze from both sides I sense a squeeze on the mind, and it comes from two directions at once. From outside, AI is getting better at thinking than we are, or at least at the kinds of thinking we're paid and praised for. Summarising, analysing, drafting, arguing a case. The very activities many of us have built our sense of worth on are becoming things a machine does faster. From inside, we're handing our thinking over voluntarily. Not because we're lazy, but because attention has become scarce and hard to win. When it's that hard to be noticed, the tool that helps you produce more, faster, starts to feel essential. So we reach for it before we've sat with the question ourselves. One side devalues our thinking. The other quietly erodes the capacity for it. What the research says, and doesn't I went looking for evidence to test this, not to confirm it. Some of it holds up surprisingly well. The clearest finding comes from a set of randomised trials by researchers at Carnegie Mellon, Oxford, MIT and UCLA, published this year. After only about ten minutes of AI help, people did worse once the AI was taken away, and they gave up more readily. The damage was concentrated in those who took the AI's answers. Those who asked only for hints were fine. A separate study published by the American Psychological Association found something subtler. Relying on AI didn't seem to reduce people's cognitive ability, but it did reduce their confidence in their own reasoning and their sense of owning their ideas. Most participants agreed that the AI had done most of the thinking. The usage data points the same way. Anthropic reported that conversations where people hand over a whole task rose from 27% to 39% in nine months. But the evidence cuts both ways, and I want to be honest about that. In a trial of nearly a thousand Turkish secondary school pupils, those given ordinary ChatGPT scored 17% worse on an exam without it. Those given a tutor version that guided rather than answered suffered no harm at all. At Harvard, a carefully designed AI tutor produced more learning in physics than a good active-learning class. When AlphaGo beat the world's best Go players, the field didn't collapse. Human moves afterwards became better and more original. On my own evidence base, social media, Reddit comments remain about 98% human. The outsourcing is concentrated in polished, real-name posting, LinkedIn above all, not in ordinary conversation. So the honest conclusion is narrower than my hunch. Using AI to get answers erodes persistence and a sense of authorship, and it does so quickly. Using it to scaffold thinking doesn't. Whether any of this adds up to a lasting change across a population, over years, nobody has tested yet. The deeper worry Here is the part that troubles me most, and it's the part with the least evidence. A sense of self shouldn't rest on the mind. But I think for many people it does. I don't see much practice of embodiment around me, certainly not in the professional world, where we're rewarded for living from the neck up. Without a felt sense of the body to come home to, the mind becomes the anchor. We are what we think, what we know, what we can argue. So what happens when that anchor is squeezed? When the thinking is both outperformed and outsourced? I suspect the cost isn't really lost thoughts. Thoughts can be borrowed. What can't be borrowed is the process: sitting with a question you can't yet answer, feeling it in the body, waiting until something forms. That's where meaning gets made, and where an internal compass gets built. Skip it often enough and you may end up with a self that curates outputs rather than one that generates them. There's some real science behind the body's role. People more aware of their own bodily signals tend to make better intuitive decisions. Traders who sense their own heartbeat more accurately have been found to perform better and last longer. And in therapy research, how deeply a client contacts their felt sense predicts whether they change. What nobody has studied is whether AI use is weakening that contact. That's my hypothesis, not a fact. A transition, not a catastrophe I don't see this as bad in the long run. I see it as a transition. If a self built on the mind loses its footing, that's also an opening. Many contemplative traditions would say identifying with our thinking was the problem all along. A crisis of the mind-made self could be the start of finding something steadier underneath. But I'm wary of assuming that's where it goes. A vacuum gets filled, and not necessarily by the body. Embodiment is the hardest path, so it may be the least likely default. The easier anchors are borrowed ones: a tribe, an ideology, a brand, outrage, or AI itself as companion and mirror. Which way it goes may depend on whether people are met when the ground shifts. That's the question I keep coming back to. An invitation rather than a warning I started out wanting to write a warning. I'm ending with a question instead, because I don't know if what I'm seeing is true, and handing you a conclusion would rather defeat the point. So here's a small experiment. Next time a question matters to you, before you ask anything or anyone, sit with it for a minute. Then ask yourself: where do I notice this in my body? If an answer comes, you still have your compass. If nothing comes, that's worth noticing too. Sources Liu, Christian, Dumbalska, Bakker and Dubey (2026), AI Assistance Reduces Persistence and Hurts Independent Performance American Psychological Association (2026), Overreliance on AI programs may undermine confidence at work Anthropic Economic Index Bastani et al. (2025, PNAS) and Kestin et al. (2025, Scientific Reports), summarised The Register (2026), Pangram study of AI content on social media Shin et al. (2023, PNAS), superhuman AI and human decision-making in Go Dunn et al. (2010, Psychological Science) and Kandasamy et al. (2016, Scientific Reports), on interoception and decision-making

Whilst waiting for 1) the simulations to build organisations from seed for the custom AI model and 2) volume testing the system with many agent, I got bored and built the Front Desk - what do you think? markaspall.github.io/the-gro…

Atoms make cells. Cells make organs. Organs make a person. Most of us stop the ladder there. But it doesn’t stop at us. Every day we each do small things. We buy this, share that, say yes to a deadline we know is unreasonable because everyone does. None of it is large. But it adds up, and what it adds up to comes back to us as the world we live in: prices, norms, the mood of a country, what feels possible. It behaves like an intelligence. One with no one home. It takes in everything we do and hands back the result, and because we don’t see it as ours, it runs in the dark. We meet it as the news, as something done to us. We almost never ask what it’s showing us about ourselves. We can’t see the whole from where we stand. A liver can’t see the body. But we can build places where the whole talks back. First an organisation, then a network of them, and in time, perhaps, something of society itself. That’s the thinking underneath The Grove. Chapter 2 of the book, written in public: markaspall.github.io/the-gro… What do you think our collective actions are showing us right now?

A live tour of a simulated organisation running on the agentic bossless organisation operating system. The new UI is really intuitive. The AI is really dialed in now. What do you think? The Grove is the first network of next wave organisations. Full details here: markaspall.github.io/the-gro…

There's a word in physiology for the sense a body has of its own condition: interoception.1 Hunger, strain, ease, the racing heart before you've named the fear. It's not one of the five senses we learn about at school, and it's arguably the one a living thing can least do without. A vehicle doesn't need it. A pilot reads dials. A living thing has to feel its own insides, or it can't look after itself. markaspall.github.io/the-gro…

If you're into living systems, you might like this : an organisation that senses, steadies, repairs and learns, and holds the whole without a centre. markaspall.github.io/the-gro…

If you're into DAOs and Web3, you might like this : every organisation a DAO from day one, with no tokens or gas, and a say that money can't buy. markaspall.github.io/the-gro…

If you're into agentic AI, you might like this : an operating system that is itself agentic, with an AI that earns its standing and never gets a vote. markaspall.github.io/the-gro…

If you're into self-management, you might like this : Holacracy, sociocracy and Teal ideas, with the structure held by the system instead of by tired people. markaspall.github.io/the-gro…

I can't get to the checkpoint! Anybody here able to? 100% WebGPU shader. Shared as Claude Artifact. claude.ai/artifact/3ebU7k7Cs…

Some of you monkeys on here might be smart enough to see what this means: ...we're training a model on organisations. The Grove is the first network of organisations running on the bossless organisational operating system. These orgs, have no bosses, and people and AI work side by side. Its structure grows from the work itself: branches with a purpose, accountabilities that answer for the work, decisions made by consent. Wise-Eye is the guide inside it. What we’re exploring now goes further. We’re giving each organisation a sense of itself. We’re training a small model on organisations as they live, day by day: their whole shape, every piece of work and where it found its home, the bets and probes, what rises in the Field, how full the Clearing is, the proposals and the tensions behind them, and what followed. It doesn’t decide anything. It gives the organisation interoception, a felt sense of its own body, reflected back to everyone in it equally. Nobody holds the whole. Everyone can ask the whole, and respond where they stand. What that could feel like: Work that can’t find a home. “Work like this keeps arriving and nothing answers for it. Organisations at this stage usually grow a branch for it.” Strain before it breaks. “The Kitchen looks the way branches often look a few weeks before they divide: work arriving faster than it closes, more of it waiting on one seat.” Drift from purpose. “Design’s purpose speaks of community groups. For six weeks nearly all its work has come from paying businesses.” A mirror, not a verdict. Seeing what a proposal would do before it’s decided. “If Workshops folds into Design, about a third of what arrives each week lands there, and Design’s open work roughly doubles.” People still decide. They just see further. Bets that remember. “This claim reads like three earlier bets that didn’t hold.” Or: “Bets probed in their first week mostly get an answer. The ones left alone mostly fade.” What the Field is sensing. Themes rising across many people’s offerings before anyone names them. Offerings that resonate but never crystallise, because something wants to happen that the shape has no room for yet. The Clearing as a vital sign. Read only at the level of the whole: gatherings thinning after a hard decision, often the first signal of a death no one could put their finger on. Tensions that keep returning. The same tension raised again and again in different words, or clustering on the boundary between two branches, where the structure is asking to change. Power gathering quietly. “Over five weeks, three seats and most of the budget have moved to one place. Decisions on money now rest on two.” Shown as shape, never as a judgement of anyone. Adapting to the world. “The work changed a month ago and the shape hasn’t followed. Organisations that followed a shift like this found their flow again within weeks.” And beyond the single organisation: A network that learns. Each organisation is sovereign, its own DAO, with its own data. What’s shared is the learning, never the records: a growing instinct for how organisations live, grow, divide and thrive, which every organisation’s own memory sits beside. Organisations finding each other. Shapes that complement each other. Work one organisation turns away that another was made for. Starter packs grown, not written. New founders begin from the patterns of organisations that bloomed from a similar spark. The rules stay the law, and consent stays how decisions are made. What changes is what an organisation is. It stops being a structure people maintain and becomes a living thing, an organism made of people and AI. It senses, it reflects, and its people respond. That loop is where collective intelligence begins.

An operating system for organisations without a boss: the design, the rationale, and how we decided. This is a book being written in the open, while the thing it describes is still being built. It isn't complete. We don't claim that what's in it is right. Some of it will turn out to be wrong, and when we learn that, we'll change it and the history here will show what changed and why. Life is an experiment, and so is this book. markaspall.github.io/the-gro…

I think it's time for a book. Here's the start of the birth of The Grove - one of the first networks of inter-operating sovereign organisations. Join me on this journey - I'd love to hear your views and insights. markaspall.github.io/the-gro…

Just some vibe coded paragliding slop