Reflections from Summer at the Wharf 2026
Making is how the thinking gets done. And now that it's possible at the speed of thinking and costs almost nothing, it stops being the cheap way to find the answer and becomes the fast way to build a stronger one.
This is obviously bigger than design. For most of human history the hard part was making the thing at all: the hands, the craft, the materials, the time. It's why a book was once a treasure, and a painted likeness the height of luxury. Every so often something new comes along to collapse the cost. The potter's wheel. The mill. Paper. The printing press. The camera. The factory. Electricity. The spreadsheet. And every time making got cheaper, scarcity simply moved upstream, to whoever chose what was worth making and could tell good stuff from bad.
Summer at the Wharf kept returning to the same liberating truth. Making stuff is now stupidly cheap. That makes us giddy. But for consultants who make, the cheapness kind of misses the point. We’ve never treated making as the step after the strategy, the bit where someone builds what the deck already decided. It is the strategy. We make to find it.
In the old days (i.e. last year) we made rough versions first, not by choice but because fidelity was expensive. That trade-off is gone. Now the making can be several versions at once, or hundreds, at full fidelity, wired into a client's real data and systems, tested at real scale, and built so the code that proves the idea is the code that ships. The prototype stops being a sketch you throw away. It's the first version of the real thing.
In our little digital corner of it, the scarce thing was still the ability to build. You needed designers, engineers, budget, months. Making was the bottleneck, so making was where the value sat. Then AI came along, and now we can make a working version of almost anything we can imagine in an afternoon. The question stops being can we make it and becomes the harder one: should we, is it any good, will anyone believe in it. What the hell should we make? The answer has never actually changed: things that solve real problems and make people's lives better than before. It's just that this answer has always been the easiest one to lose sight of. Under the effort, under the politics, under the sheer difficulty of building anything at all.
The value has emigrated from making to judgement. To discernment, taste, framing, and the political skill to get organisations on board: the soft, messy, unglamorous, human things we used to treat as the wrapping around the real work. They kind of are the work now. Or at least the bits that are scarce. And scarce is just another word for valuable.
This wasn't the headline of any single session. It was the undertow, pulling at all of them.
Taste, when everything is makeable
Made by Many's Tom Harding kicked Day One off with a properly uncomfortable question for a panel of designers: if anyone can design, what is a designer for?
The panel's answer wasn't totally reassuring. When making anything is easy, making the right thing, and recognising that when you see it, becomes almost everything.
Jade Pughe, a creative technologist at Mother London, was blunt: AI hands you the blend of every answer already on the internet, and the blend is inherently 'mid'. Joel Gethin Lewis, who teaches at UAL's Creative Computing Institute, explained why: an LLM is built to find the statistical midpoint, picking the next word that best fits the words before it. Which means the most interesting design is precisely the design that refuses to go there. Adam Morris, who leads product design at The Economist, went further: these tools make it frighteningly easy to ship plausible experiences nobody wants. Rigour still matters. Whether something is even worth building matters more.
That is taste, and taste turns out not to be decoration. It's the filter. In a world drowning in generated plausibility, the filter’s worth more than the generator.
You could watch the same truth in Iain Tait's session later that day. He shared a blizzard of beautiful, ridiculous and brilliant AI experiments. But the striking thing was never that the machine could make them. Of course it could. The striking thing was Iain's mind. The questions he's asking, the way they made us laugh slightly uncomfortably, his instincts for which experiments are worth showing. One person's thirty years of knowing what's good.
Framing, or asking the right question
Taste tells you whether a thing is good. Framing helps you decide what thing to make in the first place, and it is the rarer skill, because it happens before anything exists to react to.
Geraint Jones, who runs marketing transformation at Informa, the company behind Cannes Lions, described the past eighteen months. The rush to adopt AI doesn't always start where innovation is supposed to start, with a customer problem or a business problem. It often starts with a share-price problem: shareholders leaning on chief executives to be seen doing something with AI, and chief executives obliging. His job now is quietly unwinding it, walking leaders back through the decisions they never quite made to reach the benefits they should have started from.
That is a framing failure. Start from the wrong question and no amount of cheap, fast making will save you. You will simply arrive, very efficiently, at a beautifully prototyped answer to a question nobody even asked. Geraint put the same warning another way: you can rush to produce tooling that makes you faster, but faster is not better. Speed is now free. Direction is not.
Legitimacy, or getting anyone to say yes
Even with the right thing well made, there is the oldest scarcity of all, and the innovation panel named it more plainly than I have heard it named in a long time.
The real obstacle inside large organisations, argued Ivan Heredia, who's steered change through Disney, is not a shortage of ideas. Big organisations are drowning in ideas. The bottleneck is permission. It is not, as he said, a technology problem but a legitimacy one. The job is not having the idea. It is building belief the slow way: showing people small real things, over and over, until enough of the right people trust it to let it grow.
We tell ourselves the rare thing is the idea. It is not. Not the tool, not the pitch, not the grandiose vision. The rare thing is the person who can walk an idea through the politics, the budget rounds and the sceptical stakeholders, and wrangle all of that into something real. Innovation, as Ivan put it, is not real until an organisation actually does something. Until then it's just a bunch of slides.
The designers, in their own session, reached the same place by a different door. One described the job as shepherding the work, convincing other people it is worth doing, and holding the value of the idea from the beginning all the way through. Two sessions, two vocabularies, but the same conclusion: making was never the hard part.
Which is where cheap making earns its keep all over again. A working prototype is the most persuasive proof point there is, and it quietly changes the politics of a room. Joel, the educator on the panel, described how it used to be the person who shouted loudest, or arrived with the grandest vision, who won. Now, as he put it, if you're not turning up with a working prototype, what are you doing? The prototype flattens the hierarchy. The best idea can win instead of the loudest voice, because everyone can see and hold the same real thing. Making does not only find the answer. It earns the legitimacy to act on it.
There is an outward-facing version of the same scarcity. Sarah Gold, who’s spent a decade building her company IF around digital trust, reminded us that we’ve all by now experienced a machine confidently lying to us. Once that has happened, trust stops being a nice-to-have and becomes the thing a product lives or dies on. Inside the organisation we call it legitimacy. Outside it we call it trust. Either way. It's a finite resource: slowly earned, quickly spent, and impossible to manufacture on demand.
The pattern under the panels
Put it together and it stacks into a single, slightly unfashionable claim.
When making was expensive, you could get by being good at it. Now that making is cheap, everything left over is the whole job: knowing what is worth making, telling the good from the merely plausible, framing the real problem, telling a powerful story about it, earning the belief to act on it. None of this can be automated, because these are judgement calls, not production tasks. And judgement is stubbornly, valuably, reliably human.
You could read all this as an argument for speed. It isn't, quite. Making used to take months. Now it doesn't, and the best thing to do with the time you get back is think. My hope, and it's the optimist's version, is that as making gets cheaper, the friction worth paying for is the thinking.
We're not entirely neutral in all this. Made by Many has spent the better part of two decades in the awkward gap between strategy and execution, insisting you make your way to an answer rather than talk your way to one. We used to defend that by pointing at the making. I've come to think the making was never quite the point. The point was the judgement it forces: what to build, what to cut, what's good, what's real, what an organisation will back. Cheap tools haven't weakened that bet. They've settled it.
This was the conclusion Summer at the Wharf kept coming back to from every direction. As machines get better at making, taste and discernment and framing and trust become more valuable, not less.
The tools got cheap. Judgement got expensive.
Spend accordingly.
