Category: Economy & Society

On the morning of Thursday 27 August I opened my dashboard to see what the Americans had published overnight, and the inflation panel stayed green. Headline PCE 3.7% for July, core PCE 3.3%, both up 0.2% on the month. My threshold sits at 3.5%. Green, because the series I decided to feed that panel, back when I was building the thing, was core PCE, the number the Federal Reserve uses to set policy. Had I fed it headline PCE instead, the same release, the same morning, the panel would have gone red and I would have been working down my guardrail checklist instead of making coffee.

The gap between those two numbers is 0.4 percentage points. The decision that produced it took me about 1 minute, somewhere in a weekend of building, and I had not looked at it again until that morning.

I have no salary. I finished up in November 2024 and I now run a household on savings, dividends, interest and rent, against annual expenses that are mostly fixed and mostly not negotiable. So the panel is not an academic exercise; it is the thing that tells me whether to change what I am doing. Which makes the choice of series a real decision, and I want to be clear that I made it fast and I made it by copying somebody else.

Why the Fed leaves out the things I actually buy

Core inflation exists for a defensible reason, and I want to give it its due before I criticise my own use of it.

Food and energy prices swing hard on weather, on geopolitics, on shipping and on global commodity markets, and most of those swings reverse. Monetary policy works with lags of 12 to 24 months and often longer, so a central bank that moved the cash rate every time petrol jumped would be tightening into a shock that had already passed and easing into one that had not yet arrived. Stripping the two noisiest categories gives policymakers a cleaner read on the slower-moving pressures they can actually influence: labour costs, capacity, services demand. The formal target in both countries is still the full basket over time. Core is the working instrument, not the destination.

Picture a smoke alarm wired to ignore the kitchen, because the kitchen sets it off every time somebody browns onions and the false alarms train the household to pull the battery out. Ignoring the kitchen is the right engineering choice on 99 nights out of 100. It is the hundredth night that concerns me, because the fire usually starts in the kitchen.

What the kitchen looks like here

New Zealand’s CPI rose 4.1% in the year to the June 2026 quarter, and 1.5% in the quarter alone. That is a two-year high and the third straight quarter outside the Reserve Bank’s 1% to 3% band. Underneath it: petrol up 27.5%, other vehicle fuels and lubricants, mainly diesel, up 71.0%, electricity up around 12%, local authority rates up 8.8%.

Now the part that should stop anybody using a core measure in this country. Stats NZ published the counterfactual: take the fuel moves out and CPI rose 2.9%, comfortably inside the band. Which means the entire breach of the target, the whole of the gap between “on track” and “three quarters over”, sits in the category that core measures are designed to exclude.

Against that, wages. The Labour Cost Index came in at 2.0% for the year, average ordinary-time hourly earnings around 2.8%, median hourly earnings 2.7%. All of them well under 4.1%. Real incomes are falling for most people who work for a living, and for those of us who do not, the equivalent arithmetic is similar. The real return on anything held in cash or short deposits is negative once you use the number that reflects what you actually spend. The OCR went to 2.50% in July, its first rise after sitting at 2.25% since November 2025, and deposit rates are still nowhere near the 4.1%.

And the labour market kept doing what I said it would do. When I posted about unemployment reaching 5.6% in the June quarter I noted that I had bet on a further rise, and the 3-quarter and 8-quarter moving averages I run were the reason I was willing to say it out loud. The June print came in worse than every major bank’s pick; BNZ and Westpac had 5.4%, ANZ and ASB had 5.5%, Kiwibank had 5.3%. Underutilisation went from 12.9% to 13.8%. Rising unemployment with 4.1% inflation and a central bank that has just started raising again is an uncomfortable combination for anybody drawing down capital, and it is the reason I have panels at all.

The variable most people are not watching

Diesel, up 71% in a year.

Petrol is the price everybody sees, because it is on a sign at head height on the way to work, and when it falls people relax. Diesel is the one that moves through the system. It runs the trucks that bring food to the supermarket, the tractors and the harvesters, the coastal shipping, the generators, the excavators on every construction site, the freight that sits inside the price of a fridge before anybody puts a margin on it. A sustained rise in diesel does not stay in the transport line of the CPI; it turns up months later in groceries, in building costs, in the price of getting anything repaired.

We are a small open economy and a net importer, so a good deal of this arrives at the border rather than being made here. A stronger New Zealand dollar takes some of the sting out of it for a household like mine, at the cost of the exporters who earn the foreign income the country runs on. Sitting where I sit, drawing on savings and buying imported goods, I would take the stronger dollar. I recognise that is my position talking, and that the farmer down the road would tell you the opposite with equal justification.

What I changed

The panel now runs both series. The PCE threshold stays where it is, one number, at 3.5%, and headline PCE is what has to clear it from here, because the threshold exists to tell me when my own cost of living has moved enough to change my behaviour. Core sits alongside as context, because core tells me something different and something I still need: what the Reserve Bank and the Federal Reserve are likely to do next, which drives my deposit rates, my bond ladder and the discount rate on everything I own. One threshold, two series, two questions.

The rest of it is unglamorous. I plan the driving instead of taking trips as they occur to me. I shop differently, and I have become the sort of person who reads the catalogues and checks the online prices at Pak’nSave, Woolworths and New World before leaving the house, and who runs the fuel coupons, and who knows what a kilogram of chicken thighs costs at each of them. Every three weeks or so we drive up to Auckland to see our daughter, and the Costco run and the Asian grocery shopping are planned into that same trip, Foodie, Tai Ping, and now Stacks, which opened recently at Sylvia Park. One drive, the visit and the shopping both, because the fuel is being spent either way. I hold the fixed annual expenses under review even when nothing appears to be wrong, because rates, insurance and electricity have all gone up faster than the average and none of them respond to being negotiated with. And I watch diesel, weekly, the way other people watch the exchange rate.

This is not financial advice, but this is what I did.

The broader point is the one I would have missed if the July release had come in at 3.4% instead of 3.7%. A dashboard is only as good as the question you asked it, and the moment where you choose which series to plug in is the moment you decide whose question that is. I had unthinkingly adopted the Federal Reserve’s question, which is a good question, carefully constructed by people who understand the lags better than I do, and which is about managing an economy over two years. My question is whether the money lasts. Those are not the same question, and the number that answers one of them will quietly answer the other one wrong.

Most of us build our measurement systems by copying whoever seems to know what they are doing, and then we trust the output because we built it ourselves.

I changed the series this week. That took about 1 minute too.

I asked an AI if it could replace me. The honest answer was sitting in the two jobs I deleted.

Back in 2024 I updated my CV for a role at Open Polytechnic, and I left two things out. ANZ. ABN AMRO. Two banks, real years, real work, missing from the page.

I wasn’t job-hunting for sport. I was at Te Pūkenga while the whole thing was being unwound, and a good slice of the tertiary education sector was reapplying for its own jobs at the same time. The CV had to do real work, against a crowd, in a hurry.

I’m not ashamed of them. They just didn’t fit the story I was telling that one reader. A CV is a signal. You’re pointing at one person who has to decide something about you with very little to go on, and you choose what they get to see.

That’s the whole game. Most people never name it.

So when I sat down a few weeks ago and asked an AI the question half my generation is quietly asking, “Can you replace me,” I already suspected the interesting answer wasn’t going to be about what I know.

The part I’ll concede first

Let’s be clear about where the machine wins, because denying it is how you lose the argument before you start.

AI is better than me at most of what I used to get paid for, and faster. Drafting a strategy paper, producing three architecture options with the trade-offs laid out, writing the policy, building the dashboard, generating the risk register. These were once high-value cognitive outputs. They were what a senior person produced to prove a senior person was in the room. They are now close to free.

And here’s the uncomfortable bit. The knowledge underneath all that was never mine to begin with. It’s out there. It’s in the books, the case studies, the frameworks, the same material anyone could find if they read more and researched harder. Gladwell’s 10,000 hours, the tipping point, Lean, Agile, service management, the whole canon I spent decades absorbing. I didn’t invent it. I accumulated it. And accumulation is exactly the thing a model does better than a human.

So if my value were the knowledge, I’d already be replaced. I think a lot of experienced people sense this and don’t say it out loud, because the alternative feels like admitting the years didn’t count.

They counted. Just not for the reason most people think.

What the machine can’t see

Here’s what I kept coming back to. The AI doesn’t know me. It knows what I’ve published, what’s documented, what’s in the public record. It doesn’t know the things that aren’t written down anywhere: my taste, what I’d refuse to do, how I read a room, why I left two banks off a CV.

I studied economics at LSE and forgot most of it, but two ideas stuck, and they turn out to be the whole story.

The first is asymmetric information. The seller knows things the buyer can’t see. George Akerlof wrote about the used-car market, where buyers can’t tell a good car from a lemon, so they assume the average, so they only pay the average price, so the good cars leave the market and you’re left trading lemons. When the buyer can’t tell quality apart, quality stops being rewarded.

The second is signalling. Michael Spence’s idea, same Nobel as Akerlof. When you can’t show your real quality directly, you send a signal that’s expensive enough that a weaker player couldn’t fake it. The degree, the track record, the reputation you’d lose if you lied. The signal works precisely because it costs something to produce.

A CV is a signal. My About Me notes, the ones I feed to an AI so it writes in my voice instead of a generic one, those are a signal too. They’re me trying to make the undocumented legible. My taste, my values, the patterns I trust, written down so the machine can see what the public record can’t.

For about a week I thought that was the answer. Document the judgment, encode the taste, and the part of me that’s scarce becomes portable. Then I noticed the problem.

The signal gets cheap the moment everyone sends it

A signal only holds its value while it stays expensive to fake. That’s the entire mechanism. The day it becomes cheap, it stops separating the strong from the weak, and the market does to it exactly what it did to Akerlof’s used cars. It discounts everyone back to the average.

A documented “here is my judgment” file is cheap. I could generate a plausible one for a stranger in a minute, full of values they’ve never lived and lessons they never paid for. So could you. So when everyone has an AI producing polished strategic thinking, polished thinking stops proving anything. The mid-level manager now sounds strategic. The junior consultant now produces partner-grade decks. The signal floods, and a flooded signal is noise.

This is the part the AI’s own answer to me missed completely. It told me to “evolve from delivering outputs to architecting judgement systems,” to become a fractional this, an AI-era that. Reasonable, fluent, and indistinguishable from what it would tell anyone with a similar CV. A lemon. Confident, generic, average.

The drift toward the mean

I’ve watched this happen with my own hands, in a small and stupid way that taught me more than the economics did.

I draw a comic strip. Bald characters, no mouths, just eyes and a nose, dry organisational humour. I trained Gemini on a few years of my strips and asked what it needed to draw in my style. After some back and forth it rendered a panel that looked exactly like mine. Then the next one had mouths. And collars. And shirt buttons, on a character that wears nothing but eyes and a nose.

Gemini drifts. Left alone, it slides back toward its training data, toward the average idea of what a character should look like, and I have to reload my source strips and re-edit the prompt to drag it back. Five or six rounds for a strip I’m willing to publish.

That’s the mechanism, sitting right there in a cartoon. The machine reverts to the mean unless something private and specific keeps pulling it off course. Feed it your broken process and you get a faster broken process. Feed it the public average and it gives you back a more confident average. The tool doesn’t audit what you hand it, it multiplies it, at a speed that makes the sameness harder to notice.

So the documented signal doesn’t save you, because the documented signal is what the machine is built to average. The thing that survives is the source material, the actual strip, the actual record of what you really did, the part you keep having to reassert because the average is always pulling the other way.

The only signal left standing

Strip it back and one signal is still expensive to fake. The track record of real choices, made under real consequence, that you would defend out loud in a room full of people who were there.

Look at the choices. Lifting staff engagement 64% at Stats NZ while the organisation was mid-transformation and exhausted. Taking Priority 1 incidents at Inland Revenue from 33 in one year to 3 two years later. Pulling 20% out of platform costs at Te Pūkenga while the whole institution was being disestablished underneath me. Those aren’t artefacts a model can generate. They’re decisions I owned, with my name on them, where being wrong would have cost me something real.

And the omission belongs in that list too. Leaving ANZ and ABN AMRO off the CV was a judgment, made for a specific reader, with reasons I could explain and defend. The AI can write the CV. It cannot make the call about what to leave out, because that call needs private context it doesn’t have and a consequence it will never carry. The edited CV is the proof of the whole argument. The judgment is in the deletion, and the deletion is the part that isn’t documented anywhere.

So what’s left to learn

If the process work and the first-pass decisions go to the machine, and the polished output stops signalling competence, what’s actually worth learning?

More knowledge won’t do it. The machine already has more than you can accumulate.

Two things, I think, and neither of them is new, they’ve just become the whole job rather than the senior part of it.

The first is deciding what’s worth deciding. Framing the real question out of a mess of people, politics, money and history, before anyone produces a single artefact. The machine answers questions beautifully. It has no idea which question matters, and it never carries the cost of getting that wrong.

The second is signal literacy, in both directions. Reading which signals are still real when the cheap ones are everywhere, and being able to produce one that holds. In a market drowning in fluent, confident, average output, the scarce skill is telling the lemon from the car, and not becoming a lemon yourself.

The years didn’t buy me knowledge the machine can now match in seconds. They bought me the taste to know which ten minutes of a problem actually matter, and the scars to prove I once paid for getting it wrong.

The machine has all my knowledge and none of my consequences.

That’s the gap. I’m not worried about it closing.

I watched the internet arrive. It landed on my desk as a working problem long before it was a headline, and it rearranged how the work got done while we were still looking the other way. One year we were faxing change requests and walking floppy disks between buildings, and the next we were arguing about whether email counted as a real record, whether you could trust a document that nobody had signed. Nobody stood up and announced it. It seeped in, and by the time we noticed, the old way of working had already left the building.

Twenty something years later I am watching the same thing happen again, except the word on everyone’s lips is AI instead of internet, and the people panicking and the people cheerleading are roughly the same shape they were last time, just younger and louder and faster to post about it.

So let me say the thing this whole piece stands on. The business cycle moves in circles. We keep drawing it as a straight line, an arrow that only ever points forward, and the drawing has been wrong the whole time. Things come back around, dressed in different clothes, carrying a new acronym, and we treat each turn as if it has never happened before in the history of the world. It has. We just forgot, because in between the turns we adjusted, we adapted, we made the strange thing normal, and then somewhere along the way we mistook that normal for permanent.

When I was made redundant at 57, that was a cycle too, though it did not feel like one at the time. It felt like an ending. A career of 30 plus years in technology leadership, a last role running platform engineering with hundreds of people reporting up through me, and then redundancy, and then a very quiet Monday morning with nowhere I had to be. I have since stopped calling it retirement. I call it portfolio life, which is partly a nicer word and partly the truth, because the world did not stop turning when my old role disappeared. It just turned, the way it always does, and left me to work out where I now stood on the wheel.

So I am not predicting anything. I am useless at prediction, and so is almost everyone selling it. I am recognising something I have seen before.

The same river, a different season

In 1995 Peter Drucker published a book called Managing in a Time of Great Change. Read the title again, slowly. A time of great change. He was writing about the internet, globalisation, the early tremors of the information economy, the sense that the ground under management had shifted and the old maps no longer matched the territory. People read it then the way people read the AI commentary now, as if someone had finally found words for a feeling they could not name.

Here is what struck me when I went back to it. Almost nothing in that book was actually new in 1995. Drucker was not inventing fresh wisdom for a fresh crisis. He was reminding people of things they already knew and had stopped doing, because the long stretch of relative calm before the storm had let them get comfortable, and comfort has a way of erasing the lessons of the last storm. His advice reads as evergreen because it is. It works in 1995, it works in 2008, it works now, and it will work in whatever year the next great change arrives wearing whatever the next acronym turns out to be.

The reason it feels new each time is simple, and a little embarrassing once you see it. We adapt. We are good at adapting, it is one of the better things about us. But adaptation has a cost. Once the new technology becomes ordinary, once the new process is just how we do things now, we file away the discomfort that taught us anything, and we go back to managing the present as if the present were the permanent state of affairs. Then the wheel turns, the norm we adjusted to stops being the norm, and we stand there blinking, certain that nobody has ever been this confused before.

What actually does not change

The technology changes. The application of it spurs new processes, new business models, new centres of power. The internet gave us e-commerce and remote work and the slow hollowing of the high street. AI is busy doing its own version of that right now, and the specifics will surprise us, they always do.

But underneath the changing technology, human behaviour sits more or less still. The same fears, the same herd instincts, the same rush to either panic or worship whatever is new. The same handful of people quietly getting on with the work while everyone else argues about whether the sky is falling or ascending.

There is an old line I keep coming back to. There are three types of people in this world. Those who wait for things to happen. Those who do not even know that anything is happening. And those who make things happen. The technology of the day does not change those three categories one bit. The internet did not abolish the people who waited, and AI will not either. All a great change does is sort people more sharply into the bucket they were already heading toward.

Nobody knows what happens next, and the confident ones are usually selling something. The question you can actually answer, because it is about you, and you are the one variable you have some say over.

Drucker’s list, and why we forget it

If I strip Drucker’s advice back to its bones, the most useful piece is the first one, and the rest hang off it. Stop trying to predict the future, and start understanding the impact of what has already happened. The future is fog. But the change that has already landed is sitting right in front of you, fully formed, waiting to be read. The internet had already happened by the time most companies got serious about it. AI has already happened. The work is to metabolise what is already true and act from there, which turns out to be a different muscle than forecasting, and a much more reliable one.

The rest of his list is the same sort of unglamorous, durable common sense. Know what you are genuinely good at, and protect it, because a changing market punishes the company that forgets its own strengths and goes chasing someone else’s. Pay attention to the people who are not your customers, because the reason they stay away usually tells you more about where the world is going than the praise of the people who already love you. Treat information as the raw material it has become. Watch how new industries quietly redraw the map of who holds the power, the way the tech firms did in the 90s and the way the AI firms are doing right now, while the rest of us are still looking at where the power used to be.

None of that is clever. That is exactly the point. It is just true, and it stays true across every turn of the cycle. We do not fail to follow it because it is hard to understand. A twelve year old could understand it. We fail to follow it because between the storms we got comfortable, we adjusted to a norm, and we forgot that the norm had an expiry date stamped on it the whole time.

Standing on the wheel

I do not think AI is a revolution in the breathless sense the headlines want it to be. I think it is a bicycle. It can carry you a great deal further than your own legs, much faster, and it will not get you anywhere at all if you do not know how to ride it, and it will happily tip you into a ditch if you treat it like a horse that thinks for itself. The question was always whether you would get on and pedal, and in which direction, and that question is as old as every tool humans have ever picked up.

So when people ask me how to make sense of a world that is changing, I do not give them a forecast. I tell them to look down at the wheel they are standing on and notice that it has been here before. Read the change that has already happened. Hold on to what you are actually good at. Watch the people drifting away from you more closely than the ones cheering. And decide which of the three types of person you are going to be this time around, because the wheel does not care, it turns either way.

I have ridden two of these turns now, the internet one and this one, and I lost a career between them and found a quieter, better-fitting life on the far side. The technology was different each time. I was not. And that, more than any prediction, is the thing that has made sense to me.

It’s worked for me.