So, How Was the Expo?

I had gone to Brussels expecting a reprieve. After months in Vietnam, where the heat settles onto the factory floor like a second roof and difficult working conditions are simply the working conditions, a European summer felt like a small mercy waiting at the other end of the flight. Brussels had other ideas. The hall was boiling, everyone was sweating through their shirts, and holding a sustained conversation became an exercise in pacing yourself, refuelling, and finding a quiet bit of wall to lean against. I will admit I took a certain grim satisfaction in it, because it handed me the one argument I keep trying to make in rooms like these and rarely manage to land. For those of us who fly in for an expo, this discomfort lasts a week, perhaps two, before we go home to milder air and forget it ever happened. For the people who actually make our clothes, it is Tuesday. It is every Tuesday. And with WMO warning that El Niño is likely to develop through the summer, with most models pointing to at least a moderate and possibly strong event, that everyday reality is likely to become more unstable. The industry is not drifting towards conditions in which it can run at its best. It is drifting towards conditions in which running at all becomes the live question.

This was the first time the Textiles Recycling Expo and the Future Fabrics Expo had shared a roof, the second outing for the recycling side and, for me, a first acquaintance with both. I had missed the inaugural recycling edition last year, so much of this was new ground for me, and I came away glad to have finally seen it up close. It was, by the standards of the calendar, a relatively small affair, two days and several thousand people, which is a great deal of humanity for that much floor space and so little time. The smallness was the point. I would categorise it not as a seeing expo but as a talking one. There was no heavy machinery to gawp at, no towering stands, no theatre of scale. The booths were modest, with a few finished samples, some printed sheets explaining a technology, a placard, or a brochure. What was on offer was conversation, which is exactly what an industry as distributed as ours needs and so rarely gets in the same room at the same time. We spend most of our year conversing across screens and time zones, innovators in one corner of the world, suppliers staggered across Asia, brands scattered everywhere else, and there is a depth that simply does not survive the journey down a video call. An in person exchange, by contrast, can take a conversation somewhere a fortnight of emails never would. For me, it was mostly follow ups with people I already knew, a good number of new introductions, and a tour of the innovation on show, though I will confess that very little of it was new to my own eye.

The optimism in that room ran extremely high, and rightly so. Some of the most genuine technical ingenuity in this industry is happening right now, and every year brings more of it through the door. I find myself holding a slightly different posture from a good number of my peers on this, and I want to be careful about how I put it, because it is easily misread. I am not arguing against innovation. We need it constantly, and we will always need it. What I am arguing for is a sense of proportion about how many things we can seriously carry at once. We have thousands of brands, tens of thousands of suppliers, and now what feels like thousands of innovators, and the honest question is not whether we have enough eyes to glance at every booth. It is whether we have the manpower, the money, the time, and the patience to actually test them properly. Because innovation has never once worked the way the press release implies, with the tidy arc from idea to pilot to production in a single confident sweep. For almost everything worth having, it is a matter of years, often a decade or more. I know innovators who have been in the field for fifteen years and have still not reached anything I would call meaningful scale.

So when I asked a few people at the conference what share of the booths around us they thought would still be standing in five years, the answers came back a mixed bag, as they always do, though nobody was reckless enough to say all of them. The estimates clustered somewhere between thirty and seventy per cent, with fifty being the most common guess. On the face of it, fifty per cent is a triumph. Say there are two hundred innovations in the room and a hundred make it through, and you would take that anywhere. But the arithmetic of innovation is not symmetrical, and this is the part we skate over. A hundred winners is a fine thing for the industry. A hundred losers is a far heavier loss than the winners are a gain, because of what the loss is actually made of. It is years and sometimes decades of work, of capital, of human energy, not only from the founders but from the teams who believed in the vision and bet their careers on it, undone by any one of a hundred ordinary failures, a funding round that did not close, a pilot that went sideways, a partnership that arrived a year too late, a marketing misstep. There are a million ways for an innovation to die, and most of them have nothing to do with the quality of the science.

This is why I have come to believe that our instinct should not be to scale every innovation as quickly as possible, but to create the conditions that allow the right innovations to survive long enough to become useful. Too many are arriving at once, and the industry does not have unlimited money, technical capacity, factory bandwidth or patience to carry them all properly. I think of it almost like Earth Overshoot Day. There is only so much resource available in a given year, and once we spend beyond it, we are not creating more capacity, we are borrowing from the future and pretending the bill will not come due. This is one reason funding in the space moves so cyclically. A category becomes fashionable, capital rushes in, everyone wants exposure, and then two years later the room has moved on and the same category is described as overhyped. The cyclicality is not bad luck. It is the consequence of going all in together while the momentum is high, drawing down on future capacity until the high cannot be sustained and the correction arrives. A more pragmatic industry would give the strongest and most ready innovations the support to move first, while also finding ways to keep the others alive until the supply chain is prepared to receive them. Some technologies are genuinely impressive, but they are too early for the system around them. The missing pieces are not inside the innovation alone. They are in the infrastructure, the economics, the suppliers, the brands and the adoption pathways that have not yet caught up.

I would say something to innovators as well, and I say it with real respect, because the engineering on display in this space is often extraordinary. But one of the hardest things to internalise is that the industry’s hope for sustainability is not always the same as its operating agenda. Brands may speak fluently about reduced impact, circularity and transformation, but their actual decisions are shaped by budgets, timelines, return expectations, procurement pressure, the wider economy and the immediate needs of the supply chain. An innovation has to be tested against that reality, not only against the version of the industry we would like to build.

That means focus matters. You cannot chase twenty targets at once and expect the result to become clearer with time. There is always learning involved, and no innovation arrives fully formed, but there is a difference between asking the industry to help refine something promising and asking it to absorb too much uncertainty at once. The industry does not have endless money, endless pilots, endless patience or endless factory bandwidth to fund that education. The most important question an innovator can ask is not only whether the technology works, but whether the supply chain can realistically use it. What does the mill need. What does the dyehouse need. What does the brand need to approve it. What does the supplier have to change. What does adoption look like when production is under pressure. Build around those answers. A solution has to be supply chain beneficial and supply chain ready, or at least close enough that the remaining gaps are specific and solvable. If it is not, the better move is not to push harder for scale. It is to go back, tighten the proposition, and make the technology fit the system it is trying to change.

The investment side of the week sharpened all of this. I was glad to be invited to a session hosted by Collateral Good, a genuinely well put together affair with a handful of case studies showing how innovators are working with brands and suppliers, how the workflow actually moves, what the right support conditions look like, and where the funders come in. The room itself was the comprehensive part, with VCs, family offices, public funds, government bodies, brands, suppliers, and innovators all in one place. You hear a lot at the moment that the money is leaving the textile space. I do not entirely agree. The money is still here, but it has shifted from where it used to pool. The fibre world drew the lion’s share for the better part of a decade, several billion in all, and a good deal of that attention has migrated to the more tech flavoured corners of the industry, resale platforms, traceability tools, and digital product passports. Given the wider boom in AI and software, and given that the use case inside textiles is real, that is a reasonable place for capital to go, and resale in particular has blown up.

The caveat, however, is the one that matters most. Technology money can only take the industry so far, because the largest impacts in textiles do not sit in the digital layer. They sit in the physical one. They sit in knitting, weaving, garmenting, steam, energy systems, electrification, dyeing and finishing machinery, chemistry, and of course fibre production itself. That is where the emissions, water use, chemical burden and operational risk are concentrated. It is also the part of the industry that tends to need the most patient capital and receives too little of it.

This is why two things now feel more important than ever. The first is deeper technical due diligence. Investors already do due diligence, of course, and this is not a claim that they are careless. But textile innovation carries a level of industrial nuance that cannot always be captured through market sizing, financial projections or patent strength alone. Independent experts can help answer a different set of questions. What does the innovation actually do. Where does it fail. What is it genuinely suited for. What conditions does it depend on. How far can it be implemented without forcing the supply chain to reorganise itself around the technology. Does it solve a problem the supply chain actually has, or only one the pitch deck has defined well.

The most useful question is often not whether something is investable in principle. It is whether it is investable now. A technology can be good, important and eventually necessary, and still be too early for the system around it. In those cases, capital arriving too soon can be wasted almost as surely as capital going into something that was never going to work.

The second is something I have written about before, what I think of as a failure capital layer. The idea begins with a simple inversion. Knowing what an innovation can do is important, but knowing what it cannot do is often more valuable, because that is what reveals its real boundary. This industry has a habit of treating every promising solution as though it can be stretched across every problem. It cannot. No single solution does everything, and the sooner we understand the limits of a technology, the sooner we can stop wasting time and capital trying to force it into places where it was never going to work.

That is the purpose of failure capital. You spend deliberately to find the edges. You test the technology hard enough to understand where it breaks, not because failure is the objective, but because clarity is. Once you know the boundary, you know the true application. You know what the solution is actually good for, how much money it needs, what has to be tuned, which partners make sense, and which directions should be avoided entirely.

Take a simple example, only to show the shape of the argument. Say you have a hundred million to invest and you put ten million each into ten innovations. If only three succeed and seven fail, and each success returns ten times capital, you have three hundred million back against seventy million lost, leaving you with two hundred and thirty million. On paper, that looks perfectly good. Now imagine spending ten million first on deeper failure testing, boundary mapping and technical diligence. You now invest the remaining ninety million into nine more carefully selected innovations. If six succeed and three fail, even at a more modest five times return, you have three hundred million back against forty million spent or lost, leaving two hundred and sixty million. Even if the return falls to four times, you are still left with two hundred million and, more importantly, twice as many surviving innovations.

That last part matters more than the arithmetic. Innovation does not create value only through isolated winners. It creates value through the pathways those winners open. One successful technology can validate a category, lower the perceived risk for the next company, create supplier confidence, unlock brand adoption, and make complementary innovations more viable because a missing precondition has finally been met. A higher number of survivors across different parts of the system can compound in a way that a few large financial wins on their own cannot. We tend to optimise for the immediate return, when the deeper prize is a healthier innovation ecosystem with more technologies living long enough to make the next set possible.

All of which leads back to the same place, the one factor that decides everything and gets the least airtime, which is adaptability into the supply chain. Technical ingenuity and novelty are wonderful, but a technology that is operationally ready, that slots into the environment it is going into without a string of tweaks, has already won the hardest battle. Get that right and most of the other variables remain solvable. Get everything else right and miss that one, and the best impact in the world arrives with no adoption, which is to say no impact at all. A modest technology that the industry can take up at scale tomorrow beats a brilliant one it cannot. The dyestuff world, the corner I know best, illustrates this almost painfully. There are some genuinely gifted colour innovators out there, and extracting beautiful colour from unlikely sources deserves applause. But a great colour is perhaps a quarter of the story. The other three quarters is what a dyehouse can actually do with it, and the dyehouse is not graded on how sustainable or how bio based a colour is. The dyehouse is graded on getting orders out on time, on reproducibility batch after batch after batch, and on a very high right first time percentage, with the whole thing kept as simple as possible. Those are the metrics. And the supply chain does not want to buy one striking colour from one small company, and then another from another, juggling dozens of contracts and unknown capacities and timelines. It wants colour brought together, a few distributors, and a larger partner who has been delivering exactly that for decades. One colour is almost never enough unless the colour happens to be indigo or black, and even those have been a brutal route in. The trap is that innovators try to widen the palette to reach more brands, only to walk straight into the next demand, which is that performance has to hold across all of those colours at once. Three good colours out of six is not a foundation, it is a red flag, and a buyer will pass rather than build a programme on it.

If I had to name the single change that would do the most good across all of this, though, it is not technical at all. It is that the level of honesty in the industry has to rise. I am not accusing anyone of villainy. The pressure is structural. To raise money you have to present growth, market potential, and a clear path to returns, and so the picture gets painted in perpetual sunshine. But the sunshine obscures the very thing that would make the capital work harder. If we were more honest about the real niche, the real boundary, and the real readiness, we could size the investment correctly, stop spraying money across every direction at once, and bring down both the failure rate and the amount each innovation needs, precisely because we would no longer be testing blind. The truth is the cheapest efficiency available to us, and we keep declining to buy it.

That same reluctance shadowed the recycling floor, which had perhaps a hundred innovators on show and, encouragingly, some of the big incumbents present too, who deserve the limelight far more than they get it. The industry has been oddly asleep on its own established players. There are chemical recyclers that have been operating for the better part of a decade, with tens of thousands of tonnes of capacity, whose material has been quietly going into the products of major brands for years. And yet you will struggle to find a single announcement celebrating it. Brands adore announcing offtake agreements with shiny new entrants, but the long running relationships with the proven incumbents stay curiously unspoken. I have my theories about why, and I will save them for a piece of their own, but it is a genuine head scratcher and worth the industry’s attention. If those technologies exist, are producing at real volumes, and are already trusted by brands, the obvious move is to double down and duplicate the plants. The engineering is settled, the feedstock and the awkward operational corners have largely been worked out, and the capacity is demonstrably there. So why not simply build more of what works and let the established producers who are already ahead keep scaling, rather than staking the whole future on a fleet of first of a kind facilities yet to prove themselves?

Which brings me to a word we use far too loosely, scale. A new plant at thirty, fifty, or seventy thousand tonnes gets described as being at scale, and for some materials that is fair. The entire lyocell market runs at perhaps four hundred thousand tonnes a year, so a recycler producing thirty or forty thousand tonnes is already near a ten per cent share, which is a serious number. But polyester lives in the tens of millions of tonnes, and against that a fifty thousand tonne plant is not scale in any meaningful sense. I would put it plainly. For polyester, anything below a million tonnes is not scale. If you cannot account for even one per cent, sometimes not even half of one per cent, you are not at scale, you are a substantial facility doing important work in progress.

The European recycling story leans heavily on this misreading of scale, and on a second assumption that deserves far more scrutiny, which is that Europe is necessarily the right place to build. The case for Europe is familiar enough. The regulation is stronger, the public funding landscape is more active, grants are more visible, and the policy direction appears to favour recycled content. All of that matters. But once the conversation moves from policy ambition to industrial reality, the questions become much less comfortable.

Most chemical recyclers do not produce finished polyester. They produce monomers, which then have to be polymerised back into PET before they can become fibre again. That leaves two routes. Either the recycler builds its own polymerisation capacity, which is an enormous capital undertaking in a continent that has been losing polymerisation capacity precisely because it struggles to compete with Asian production costs, or it sells those monomers to someone else. That second route sounds simpler until you ask who the buyer is meant to be.

Europe’s textile polyester production base is far smaller than its PET packaging system, and the relevant fibre grade capacity is only a fraction of the total European PET market. Even where suitable capacity exists, the headroom is finite, and PET production in Europe remains structurally more expensive because of energy, labour, operating costs and capital recovery. Smaller recyclers may well find a home for their monomers in the near term, particularly while volumes remain modest and buyers are willing to experiment. But the economics become harder as they grow, not easier, because the bigger the monomer output becomes, the more dependent it is on a downstream polymerisation system Europe does not have at the scale or cost base required.

Selling those monomers abroad brings its own problem. China is the obvious destination on paper, because it is the centre of global polyester production. It is also the least obvious destination in practice, because there is little reason for Chinese producers to pay for expensive European monomers unless regulation or customer demand forces them to do so in a way that is both strict and durable. The history of monoethylene glycol is instructive here. China has previously relied on imported MEG, including from the Middle East, but it has also pushed hard to localise supply, even through the coal to MEG route, despite the environmental burden of that pathway. A production system willing to use a more damaging domestic route to reduce dependence on imported monomers is unlikely to become a natural buyer of higher cost European recycled monomers at commodity scale.

That is the uncomfortable gap in the European chemical recycling narrative. The demand everyone points to is demand for recycled polyester, but many of the new facilities are producing an intermediate product that still needs to find a polymerisation home. Europe may be able to make the monomer, but it does not automatically follow that Europe can turn that monomer into fibre at the cost, volume and speed the textile market requires. The supply gap may be real, but it may also be sitting at the wrong point in the chain.

The regulatory confidence deserves the same scrutiny. Bottles and packaging have had nearly a decade of the very tools we are now pinning our textile hopes on, extended producer responsibility, deposit schemes, collection systems, and mandated recycled content, with collection rates that are high in several countries and around half on an EU average, and a far simpler feedstock than any garment. And still European packaging recycling has not scaled the way the theory promised. Europe lost something like a million tonnes of recycling capacity in the past 2 years, because the demand was met more cheaply by Asian output. Now we are bringing broadly the same regulatory toolkit to textiles, a much harder feedstock moving through a much harder pathway, and expecting not the same outcome but a better one. Einstein is often credited with saying that the definition of insanity is doing the same thing over and over again and expecting different results. Whether or not he actually said it, the warning feels useful here. I do wonder, genuinely, what I am missing, and what is materially different this time in the textile space that warrants such optimism.

Even if we assume that this regulatory magic somehow works out, the real bottleneck sits even earlier, in collection and sorting, where the feedstock is either made or lost. Refashion, the French EPR system, has run for about two decades, absorbed a great deal of money, and still collects somewhere below forty per cent, and this year it has been under enough strain to start closing collection points, with operators openly saying the system does not return enough to fund the work. Sorting is further along, with some excellent companies bringing NIR spectroscopy and automation to a task that is punishing to do by hand, whether by expensive labour in Europe or by overburdened labour in Asia. But almost all of that sorting reads the material, the fibre type, the colour, wearable against non wearable, dirt and contamination, and almost none of it reads the chemistry. And chemistry is where recycling actually lives or dies. A process can have polyester cleanly separated and still choke, because a finish poisons the catalyst or fouls the reactor and the output is ruined. With post industrial waste, this is manageable, because we have largely phased out the worst chemistries and we know what went on the fabric during production. Post consumer waste is where the industry is still running largely blind. These are garments that may have been circulating for ten or twenty years, carrying chemistries that are now restricted, banned, phased out or simply no longer common in modern production. At small volumes, that uncertainty can be managed through direct testing and batch level sampling. You can inspect a limited stream, test enough of it to form a view, and make an acceptance decision. But at hundreds of thousands or millions of garments, that approach starts to break down. You are no longer just sampling a manageable batch. You are making risk based judgements across messy, mixed and poorly documented waste streams, estimating the likelihood that a given stream is safe enough based on source, garment type, treatment history and whatever testing can realistically be done. You may know, for instance, that an outdoor jacket is more likely to carry complex finishing chemistry than a plain cotton shirt, but post consumer waste rarely arrives in categories that clean.

The problem becomes even harder depending on how the material reaches the recycler. If whole garments arrive, there is at least a chance to inspect, sort and test them with some intention. If the recycler receives pre shredded textiles, much of that visibility has already been lost. What was in the material before it was shredded. Who handled it. Was any chemical compliance screening done. What finishes, coatings or contaminants are now hidden in the mix. Each unknown adds another layer of risk, and very often another preprocessing step, another cost, and another opportunity for the recycling economics to weaken before the process has even properly begun.

This is why the mono material gospel frustrates me. A mono material garment carrying finishes, coatings or treatments that disrupt the recycling process may still be difficult, or even impossible, to recycle. A more complex garment, by contrast, may be far more recyclable if its chemistry is known, compatible and manageable within the chosen process. There is no clean physical rule for design for recyclability, because the deciding factor is not only the material used but also the chemistry applied, and the chemical layer is sadly not the one leading the conversation. There are good people building it (ex. Around Systems), working on systems that let brands understand in advance whether the chemistry they plan to use will be compatible with recyclers downstream. While such solutions are emerging, the pace remains slow, because the bigger problem is what the industry is ready to admit. Everyone on stage has cracked recycling and solved the elephant in the room. Off stage, in the quieter conversations, the same people are candid about the chemistries they cannot handle, and the problems they cannot yet solve. The distance between those two registers is what we need to close, and it will not close until we let the off stage honesty onto the stage.

Sitting underneath the entire week was a piece of contrasting research that I think reorders a good deal of our strategy, the new Textile Exchange life cycle assessment on virgin polyester, built on data from PET production in China rather than the European proxies we have leaned on for years. The old reference points were not merely dated, they were wrong in a consistent direction. We began with the PlasticsEurope figure of around 2.2 kilograms of CO2 per kilogram of PET. The EcoInvent update, once it accounted for upstream methane, lifted that to roughly 3.12 kilograms of CO2. But the deeper flaw was that none of it reflected where most of the world’s PET is actually made. This new assessment puts virgin PET at 4.5 kilograms of CO2 and beyond depending on the scenario, and it is still the conservative read, because it assumes some biomass energy and steers clear of the coal to MEG route that is among the most common and most polluting in China. Take the real coal pathway and the figure climbs past five. What that implies cuts two ways. For recyclers, it is almost a gift, since most of them have spent years straining to beat 2.2, found some relief at 3.12, and will now most likely be under the virgin benchmark, by a margin that depends on their grid and their efficiency and their failure rate but a margin nonetheless. For everyone else, it is a reckoning, because it means we have been working from a baseline that flatters us. Most brands set their emissions baselines somewhere between 2019 and 2023 on the lower polyester figures, which means their Scope 3 baselines, especially for the polyester heavy among them, are understated. As this data works into the standard datasets, those baselines have to be revised upward, and a target of thirty or forty five or fifty per cent reduction has to climb with them, because the starting point was wrong and the progress was never as large as it looked. The reduction already achieved is real and does not vanish, but it is now a smaller step up a taller staircase, and across the industry, measured in millions of tonnes of CO2 against a 1.5 degree pathway, the fight is materially bigger than we thought.

You cannot answer a bigger fight with a ten or fifteen or twenty per cent recycled content share. The clock on the solution providers has just sped up, and it is no longer enough to deliver. You have to deliver at scale and at a price that competes, with brands actually adopting the material rather than admiring it, and all of that on a timeline that the physics of plant building does not respect. You do not bring meaningful new fibre capacity online in three or six months. Which returns me, by a route I did not choose but cannot avoid, to the thesis I have been pushing for as long as I have been writing, that supply chain decarbonisation is the more urgent lever, and that the new polyester number has only made it more urgent. Pulling the industry’s footprint down on materials alone is a mammoth undertaking against a baseline that just rose, and it depends on a market responding faster than markets do and on supply that does not yet exist. Supply chain decarbonisation does not have that problem. The solutions exist and are available to install now. Within a year or two you can adopt almost all of them, not at full capacity overnight, but you can start immediately, place an order tomorrow, have the machine commissioned in short order, and begin banking the savings, which matter more now than they ever did and matter most of all to the polyester heavy. And even a company that uses little polyester is not exempt, because the industry’s number is what counts, not the individual’s. You can run the cleanest house on the street and still suffer if the street is on fire, because global emissions do not respect corporate boundaries. With heat stress mounting, with the worker safety reports stacking up, and with whole producing countries facing the loss of exports and labour as conditions on the floor deteriorate, protecting this year’s revenue cannot be allowed to cost the existence of future revenue. If a business wants to still be a business in twenty years, the investment is not optional, and it is not later. It is now.

The Future Fabrics side of the floor was the more diverse and, in its way, the more romantic, mycelium leathers, bio based dyes and chemistries, biopolymers, cellulosics, a little of everything, and some lovely work among it. I take it, though, not with a pinch of salt but a bucket, because the precedent is hard to ignore. We have seen a striking run of bankruptcies over the past decade, a whole next generation landscape that took the stage and has not, by any honest measure, delivered. More than three billion dollars has gone into next gen materials, and the market share is still under one per cent. I want patience and perseverance as much as anyone, and I want every one of these innovators to make it, but precedent is information, and the industry does not have the luxury of carrying everyone indefinitely in the hope that all of them come good. We need the right solutions first, which is not the same as abandoning the rest, and it is why I keep returning to buying innovations time rather than forcing them up the scaling curve before they are ready. We do need solutions now. What we need precisely is the right solution at the right time in the right order, not any solution at any time in any order. You cannot pour the ground floor and the fifth floor of a building at once. There is a sequence, and the sequence is not a bureaucratic inconvenience, it is the physics of the thing.

The analogy I keep coming back to is the body. A patient arrives at the hospital with a heart attack and, as it happens, a fractured arm and a fractured leg. No competent doctor sets the fractures first and gets to the heart afterwards. You stop the heart failing, and only once the patient is alive do you turn to the limbs, because a perfectly set arm is worth nothing on someone who died while you were setting it. The industry is that patient. The fractures are real, and waste is a genuine fracture that we must and will attend to, but there are other ways to manage waste in the meantime, and if we leave it for a while it will still be sitting there in twenty years, exactly as broken, waiting to be fixed. Carbon is the heart attack. Leave it unchecked for twenty years and there is no coming back to it, no recycling it out of the atmosphere on a Tuesday afternoon. So we have to be honest about what is critical to the mission and build on it first. The single point of failure for this industry is its carbon emissions, and everything else, important as it is, is a fracture we can set once the patient is breathing. Spend the decade on the fractures and ignore the heart, and we may end up with a beautifully circular industry that no longer has a planet to be circular on.

For all my questions and reservations, it was a genuinely valuable week, full of familiar faces, new introductions, sharp panels and the kind of unguarded exchanges that only happen when people have finally stepped away from the screen and into the same room. I will do my best to be there again next year. There is far too much from the event to fit into one piece, so I will be doing a few deeper dives in the coming days into specific conversations, technologies and questions that stayed with me after Brussels. In the meantime, to every innovator who showed up with something built, tested and ready to be questioned, I am rooting for you. And if there is any way I can be useful, you know where to find me.