A Bit of Chaos You Choose

A few days ago I wrote that confusion is the correct response to what is happening in AI, and that the disorientation people feel is shared by the leading experts in the field. That piece was about a condition arriving from outside, produced by signals landing on top of each other faster than anyone can sort them. This one is about the opposite arrangement, where the chaos is mine, I built it deliberately, and I decide how much of it comes into my life.

Two Kinds of Confusion

Most people meeting AI are confused in the first way. They receive too many messages, they are happy enough to follow whatever the mainstream press tells them, and what it mostly tells them is to be afraid. That confusion comes out of passivity and it produces anxiety, because nothing in it asks anything of the person feeling it.

The second kind comes from experimenting and experiencing while you are not yet experienced. It has the same mechanism underneath. Confusion is the inability to classify the feedback the world is giving you and to act on it, and experience is what removes that inability: through muscle memory or mental habit you assess a situation instantly and you know what to do. A cook who has made a sauce two hundred times does not think about whether it has split, they simply see it. The difference is entirely in how you arrived there. One is what happens when the world sorts you. The other is what happens when you go and find the edge of your own competence on purpose, because you know that standing at it will benefit you.

I had been practising the second kind for years before I had any way of recognizing what it was.

A Dozen Plastic Cups

When my children were small, five to eight or so, we would go to the park in summer with a dozen plastic cups. We would set them down along the path, take one each, fill them at the drinking faucet and start a water fight, the three of us, in the middle of everyone else using the park for its ordinary purposes.

People got splashed. Children, and adults too. Most often it happened behind their backs, so they would turn around not knowing who or what had hit them, and at that moment, rather than explaining, we would pull a face and point at the cups still sitting on the path, without any other explanation. A large number of them picked one up and joined in.

Watching adults who had left the house that morning with no such plan in mind get thoroughly wet and laugh about it was the best possible return on a dozen plastic cups. What made it work was not the water. It was that the faucet is for drinking and the path is for walking, and in about four seconds a set of assumed behaviours had been dislodged and replaced with a better alternative.

One Chopstick

The simpler version happens at my own table. When we have Chinese food, sometimes I set the table and give everyone a single chopstick.

They sit down and ask what they are supposed to do with it. That is a fair question, and I do not have the answer. You could pair up: two people, two plates, two chopsticks between them, and now dinner is a negotiation. You could do something else I have not thought of. The point of the exercise is the four or five seconds before any of that, when a person has to introspect fast and find a new point of balance. It is a small cognitive jujitsu move. You unbalance someone, and they either recover in a new position or they go over, and going over teaches them something too.

Why It Only Works Twice

I do not do this often, and the reason is structural. Once it is expected, it stops working. A guest who arrives knowing there will be one chopstick has already done the sorting in advance, which means the muscle I was hoping to make them use never engages. The same is true of the park: a family known for water fights is a family people walk around.

Anything that trains you by surprising you expires the moment it becomes a routine. This is the constraint that keeps deliberate confusion useful, and it is also why you cannot buy it. You have to keep inventing the next one.

Walking Is a Sequence of Unstable Events

What I want is for people to exercise their confusion the way the strangers in the park did. They did not run, they did not call the police, and most of them did not get angry. They were energised, activated, engaged, and they had fun.

Anyone looking at AI and the possibilities it represents can do the same thing with the confusion it produces in them. Embrace it, understand it, and go and experiment. What you should not expect is that the confusion will dissipate, because the nature of the technology and of this period means it will not. New stimulation, new features and capabilities, will maintain the level of surprise and unexpectedness necessary.

A toddler who has learned to walk has not acquired the ability to manage instability. Walking is a sequence of completely unstable events, a controlled fall caught each time by the next step, and it never stops being one. The effort remains, it simply becomes invisible: still muscular, effortless only in terms of concentration. Training yourself to engage with your own confusion by generating a bit of purposeful chaos works the same way. You do not reach an impossible steadiness. You become able to walk confidently through a world that keeps moving, while the people who never practised stand still, unable to read what is happening to them.

What It Costs to Try

There is an assumption that this belongs to the young, who act outside their comfort zone because they have not yet built one. Past some line we call adulthood, risk-taking is supposed to diminish.

One of the real achievements of modern societies is that we can now construct conditions in which we can afford risks that would once have been unconscionable. Consider a medieval couple opening a tavern. If it failed, he went to debtors’ prison and died there within months, and she ended up on the street with the children. There was no mechanism for separating yourself from the failure of your own initiative, and as late as 1905 there were still more than eleven thousand debtors in English prisons.

Limited liability is the invention that changed this, and it unlocked risk-taking for enormous numbers of people who could never have afforded it before. There are now places where having tried and failed is a badge of honour and nobody holds it against you.

Not everywhere. In Germany a five-year bar on serving as a managing director does exist, but it attaches to a conviction for an insolvency offence, for filing too late, for failing to keep the books, for fraud or breach of trust. Failing without fault carries no such bar. The penalty that falls on the person who tried is cultural: succeed and a certain kind of envy finds you, fail and you are laughed at, and failing badly is the worst of the two. A society that arranges its feelings this way is asking its members to stay where they are.

The legal machinery for affording risk has been built. What remains uneven is whether the people around you will let you use it.

The Invitation

So this is less a demonstration than an invitation. Find the version of small chaos that suits your household and your week. Put out more cups than there are people. Give someone one chopstick. Open the tool you have been told to be frightened of and spend an evening being bad at it, on purpose, where nothing is riding on the outcome.

You will be confused, and that is the entire point, because the confusion you walked into deliberately is the only kind you can practise with. We are all in the same boat on this, and anyone can be confidently confused in public, especially after having trained themselves on a bit of chaos. I recommend it.