This week I was in Melbourne for the AI Engineer Conference - 800+ software engineers, leaders and founders, all there to learn the latest in AI. Between the great keynotes and the hallway chats, there was something else you could feel in the room: a collective identity crisis.
For a lot of people there, the way they work has fundamentally changed. Many of the software engineers I talked to hadn't written a single line of code this year. If you've spent your whole career being the person who writes the code - the one who can do the thing most people can't - and suddenly you're not really the one doing it, that's a strange place to be.
I felt it too. I'm a designer, not an engineer, but the shift's the same: I'm not nudging assets pixel by pixel anymore, I'm further back, directing. And the doing - the part that took years to get good at - is the part getting handed off.
Which leaves a pretty uncomfortable question, the one everyone was sitting with: where did our value go?
The surface and what's underneath
The doing always had value. It was hard, it was scarce, and it's what you got paid for. But it was never the whole story. The doing is only ever worth something so far as it's grounded in something deeper.
Interactive
What's underneath
The output is the bit you can see. But what makes it good sits underneath.
What made the code, or the design, or the writing worth anything was always underneath - the expertise about what problem you were actually solving, the taste that knew when something was slightly off, the judgement that told you what to leave out. Take that away and there's nothing holding it up. There's already a word we came up with for that: slop. Doing grounded in nothing - no expertise, no judgement, no choice, no vision. And because nothing's holding it up, it's worth nothing.
So where did our value go? Where it's always been - under the surface.
Jeremy Howard talked about this in his Thursday keynote. He laid out two states you can find yourself in: positive flow, and junk flow.
Interactive
Positive flow, or junk flow
Both are flow. Both can feel good. But one makes you better and the other just feels like it does. Open each to see the difference.
Positive flow is the real thing. You're stretched, drawing on what you know, bumping up against what you don't, actually getting better. Junk flow feels the same - the sensation of being productive, of creating - but it's closer to a slot machine: your brain's switched off, you're just consuming output. It's the old split between eudaimonia and hedonia - flourishing, or easy pleasure - playing out at your desk. Hand the doing to AI without bringing your judgement, and it's easy to slide from one to the other without noticing. In other words, it's up to you to bring your value to AI.
Which is the whole point. The move is to double down on what's underneath. That value was always there - most of us just never had to name it, because the doing kept us busy enough. So get deliberately good at the parts you can't outsource: the judgement, the expertise, the direction, the call on what's even worth doing. And maybe, loosen your grip on the doing.
The tightrope
Geoff Huntley - one of my favourites from the two days - talked about the process of letting AI change the way you work. Integrating AI isn't a switch you flip. It's a tightrope you have to cross.
Interactive
Walk the tightrope
Everyone has to walk it. Nobody gets to skip the scary bit in the middle. Take it one step at a time.
Hierarchy to Intelligence
The same rope runs through the org chart. Work's been built in tall layers - the people closest to the work at the bottom, what they know travelling up through meetings and managers before someone senior makes the call and sends it back down. Every hop costs time and fragments the information on the way.
Jack Dorsey's bet - laid out in From Hierarchy to Intelligence - is that small, tight teams sitting right next to the work, with AI, can move faster than anything routed up and down a tall org chart. Proximity to the work becomes the advantage. His company, Block, is pulling decisions closer to the work, deliberately and loudly. Most companies are moving slower, but I think every company is going to have to make the shift. Geoff asked in his keynote: how long does it take a company of 6,800 people to turn itself around? The answer might be years. That's the difference between an oil tanker and a speedboat, and the AI-native startups aren't waiting. It doesn't mean the tankers are sunk. It just means the turn's a long one, and it's better started early.
So that was the week. A lot of clever people, a shared identity crisis, and a realisation about where the value always was.
Where it goes from here, nobody knows. But wherever you or your organisation are on the rope, the move's the same: take the next step.